cowork-harness 0.25.0 → 0.26.0

package/.claude/skills/cowork-harness/SKILL.md

@@ -3,8 +3,8 @@ name: cowork-harness
 description: Test or debug a Claude Code skill/plugin under Claude Cowork's runtime — sandboxed agent, default-deny egress, the can_use_tool permission/question protocol — using the cowork-harness CLI. Use when validating or regression-testing a skill, authoring or debugging a scenario YAML (prompt + scripted answers + assert:), choosing a fidelity tier, scripting AskUserQuestion / tool-permission answers, or asserting artifacts, egress, or sub-agent dispatch. Especially when a harness run no-ops an assertion, fails on an unanswered gate, false-greens, a steered answer never reaches the model, or a web_fetch is unexpectedly denied or gated. NOT for generic unit testing (pytest/vitest of your own scripts) or non-Cowork CI. Covers the skill / run / chat / record / replay / trace / decide / assertions / scaffold commands and the session-vs-scenario split.
 metadata:
   author: cowork-harness
-  version: 0.25.0
-  tracks-harness: cowork-harness 0.25.0 (baseline desktop-1.18286.0)
+  version: 0.26.0
+  tracks-harness: cowork-harness 0.26.0 (baseline desktop-1.18286.0)
 ---
 
 # cowork-harness
@@ -22,7 +22,7 @@ flagged with a loud `::warning::`, not silent — auto-answer a gate, observe an
 allowlist). This skill exists mostly to keep you out of those traps — the Gotchas section below is
 the highest-value part. Read it.
 
-> **Version note:** the facts and `file:line` pointers here track `cowork-harness 0.25.0` (baseline
+> **Version note:** the facts and `file:line` pointers here track `cowork-harness 0.26.0` (baseline
 > `desktop-1.18286.0`). If your checkout is newer, prefer the live `--help`, `SPEC.md`, and
 > `docs/*.md` over this snapshot, and re-run the bundled linter.
 
@@ -38,7 +38,7 @@ cowork-harness skill ./my-skill "do X"      # run the skill once against the sta
 Before the first command, confirm the CLI is reachable and **fail loud** (never fake a pass) when a tier's dependencies are missing:
 
 - **One-shot check.** Run `cowork-harness doctor [--tier <tier>]` first — a read-only prerequisite check that inspects Docker, the staged agent, the token, and the baseline in one pass. The bullets below explain each thing it checks (and how to fix it).
-- **CLI on PATH, recent enough?** Run `cowork-harness --version` — this skill needs **≥ 0.25.0**. If it's missing or older, prefix every command with the version floor `npx "cowork-harness@>=0.25.0" <cmd>` (Node ≥ 20), or install once with `npm i -g "cowork-harness@>=0.25.0"`. **Pin `@>=0.25.0`, never `@latest`** — `@latest` can silently fetch an older CLI and the new commands fail as "unknown command", whereas the floor **fails loud** if no compatible version is published. (≥ 0.25.0 is what gates the commands/assertions this skill teaches: `assertions --list`, `scaffold <run-id>`, `trace --view dispatches`, `artifact_json` incl. the `in:` operator, `verify-cassettes`, batch `record <dir>`/`--rerecord-stale`, `record --concurrency <N>`, record-time redaction, multiSelect/`answer:`, `verify-run` answer-coverage, `record --max-artifact-bytes`, live record-time deciders, `verify-cassettes --allow-domain`/`--allow-email`/`--allow-path`/`--allow-patterns-file` (`path` — local absolute filesystem paths — is the scanner's 4th class, new in 0.21.0; `--allow-patterns-file <path>` is a FILE of patterns, one regex per line — not a path to allow, unlike `--allow <regex>`), scenario `skills:` staleness scoping with `COWORK_HARNESS_AGENT_SCOPE=skill`, `chat --plugin`, `/help` in the REPL, `hostloop`'s native host/VM process split with its `allow_host_writes:` consent field, and `computer_links_resolve` (new in 0.22.0).)
+- **CLI on PATH, recent enough?** Run `cowork-harness --version` — this skill needs **≥ 0.26.0**. If it's missing or older, prefix every command with the version floor `npx "cowork-harness@>=0.26.0" <cmd>` (Node ≥ 20), or install once with `npm i -g "cowork-harness@>=0.26.0"`. **Pin `@>=0.26.0`, never `@latest`** — `@latest` can silently fetch an older CLI and the new commands fail as "unknown command", whereas the floor **fails loud** if no compatible version is published. (≥ 0.26.0 is what gates the commands/assertions this skill teaches: `assertions --list`, `scaffold <run-id>`, `trace --view dispatches`, `artifact_json` incl. the `in:` operator, `verify-cassettes`, batch `record <dir>`/`--rerecord-stale`, `record --concurrency <N>`, record-time redaction, multiSelect/`answer:`, `verify-run` answer-coverage, `record --max-artifact-bytes`, live record-time deciders, `verify-cassettes --allow-domain`/`--allow-email`/`--allow-path`/`--allow-patterns-file` (`path` — local absolute filesystem paths — is the scanner's 4th class, new in 0.21.0; `--allow-patterns-file <path>` is a FILE of patterns, one regex per line — not a path to allow, unlike `--allow <regex>`), scenario `skills:` staleness scoping with `COWORK_HARNESS_AGENT_SCOPE=skill`, `chat --plugin`, `/help` in the REPL, `hostloop`'s native host/VM process split with its `allow_host_writes:` consent field, and `computer_links_resolve` (new in 0.22.0).)
 - **Agent binary (sandboxed live tiers — `container`/`microvm`/`hostloop`/`cowork`).** The staged Claude Code agent is **bind-mounted** from a local Claude Desktop install, or point `COWORK_AGENT_BINARY` at a `claude-code-vm/<ver>/claude` ELF. Nothing is bundled. `protocol` (L0) and `replay` need no staged agent; for the sandboxed tiers, no agent → no run; report that, don't skip silently.
 - **Docker / Lima.** Only `--fidelity protocol` (L0) runs without them. `container` / `microvm` / `hostloop` / `cowork` need Docker (Lima for L2). If they're absent, drop to `--fidelity protocol` and **say so** — a green that never exercised the sandbox is not a sandbox pass.
 - **Auth.** `CLAUDE_CODE_OAUTH_TOKEN` (preferred) or `ANTHROPIC_API_KEY`, via env or `.env`. Minting an OAuth token needs the **`claude` CLI** (`npm i -g @anthropic-ai/claude-code`, then `claude setup-token`).

package/.claude/skills/cowork-harness/references/ci-recipe.md

@@ -1,6 +1,6 @@
 # CI recipe — replay vs live lanes
 
-Self-contained reference. Tracks `cowork-harness 0.25.0` (baseline `desktop-1.18286.0`).
+Self-contained reference. Tracks `cowork-harness 0.26.0` (baseline `desktop-1.18286.0`).
 
 **Fastest path: the packaged Action.** One step gets you `replay`/`lint`/`verify-cassettes` plus a PR
 job-summary reporter (verdict table, staleness findings, cost/turns when available):
@@ -53,7 +53,7 @@ sha256-*checked* but not hard-blocking on mismatch — it's advisory for an inte
 GitHub-hosted runners, no token/Docker/agent:
 
 ```yaml
-- run: npm i -g "cowork-harness@>=0.25.0"
+- run: npm i -g "cowork-harness@>=0.26.0"
 - run: cowork-harness lint scenarios/*.yaml          # no silent false-greens
 - run: cowork-harness verify-cassettes cassettes/    # privacy + staleness
 - run: cowork-harness replay cassettes/              # token-free content/structure
@@ -187,7 +187,7 @@ jobs:
         with: { node-version: '20' }
       - uses: actions/setup-python@v5
         with: { python-version: '3.x' }                                       # python3 only — PyYAML is bundled with the linter
-      - run: npm i -g "cowork-harness@>=0.25.0"
+      - run: npm i -g "cowork-harness@>=0.26.0"
       - run: cowork-harness lint scenarios/*.yaml                              # no-silent-false-green (needs python3; PyYAML bundled)
       - run: cowork-harness verify-cassettes cassettes/ --output-format json   # privacy + staleness gate
       - run: cowork-harness replay cassettes/ --output-format json             # token-free content/structure
@@ -216,7 +216,7 @@ jobs:
             echo "live=true" >> "$GITHUB_OUTPUT"
           fi
       - if: steps.guard.outputs.live == 'true'
-        run: npm i -g "cowork-harness@>=0.25.0"
+        run: npm i -g "cowork-harness@>=0.26.0"
       - if: steps.guard.outputs.live == 'true'
         run: cowork-harness run scenarios/ --output-format json
         env:

package/.claude/skills/cowork-harness/references/fidelity-and-answers.md

@@ -1,6 +1,6 @@
 # Fidelity tiers & answer paths
 
-Self-contained reference. Tracks `cowork-harness 0.25.0` (baseline `desktop-1.18286.0`).
+Self-contained reference. Tracks `cowork-harness 0.26.0` (baseline `desktop-1.18286.0`).
 
 ## Fidelity tiers (`fidelity:` in the scenario)
 

package/.claude/skills/cowork-harness/references/scenario-schema.md

@@ -1,6 +1,6 @@
 # Scenario & session schema, assertion catalog, web_fetch, full gotchas
 
-Self-contained reference for authoring `cowork-harness` scenarios. Tracks `cowork-harness 0.25.0`
+Self-contained reference for authoring `cowork-harness` scenarios. Tracks `cowork-harness 0.26.0`
 (baseline `desktop-1.18286.0`). If your checkout is newer, prefer the live `docs/scenario.md`,
 `docs/session.md`, and `SPEC.md`.
 

package/CHANGELOG.md

@@ -6,6 +6,46 @@ All notable changes to this project are documented here. The format is based on
 
 ## [Unreleased]
 
+## [0.26.0] — 2026-07-05
+
+Cassette/run-result **format-freeze** pass before 1.0 — fixes the parts of these schemas that become
+irreversible once 1.0's semver contract freezes them.
+
+> **Upgrade notes.**
+> - **Staleness-hash framing changed → `cassetteVersion` 8.** A pre-v8 cassette's staleness fingerprint
+>   is non-comparable with v8, so `verify-cassettes`/`rehash` route a pre-v8 cassette to **re-record**
+>   with an honest "algorithm changed" message (not a misleading "content changed"). Re-record your
+>   committed cassettes. (Cassettes without a `skillHash` are unaffected and keep replaying.)
+> - **`RunResult.unanswered` renamed to `nonReproducibleAnswers`.** Consumers reading
+>   `result.unanswered` must switch to `result.nonReproducibleAnswers` (the field never held literally
+>   unanswered gates — it holds gates answered by a non-reproducible source).
+> - **Cassette-level `readonlyFolderRoots` removed** in favor of per-entry
+>   `ManifestEntry.truncationReason`. No consumer action unless you parsed the cassette directly.
+
+### Fixed
+
+- **Two staleness-hash framing collisions (a false-negative — the risk the gate exists to prevent).**
+  `skillHash` folded raw file content after a `F:<path>\0` marker, so a file whose content embedded that
+  marker could hash identically to a two-file tree; it now folds the fixed-length content SHA
+  (self-delimiting, charset disjoint from the `F:`/`L:` prefixes). `contentSig` lacked a type marker and
+  used `:` between path and sha, so a file named `a:lnk` could collide with a symlink `a`; entries are now
+  type-prefixed (`F:`/`L:`) and NUL-framed, and the `L:` link line NUL-separates path from target. Only
+  fixable before the hash algorithm freezes at 1.0.
+
+### Changed
+
+- **Cassette format v7 → v8** (`CONTENTSIG_ALGO` 3 → 4). See the framing fix above and the upgrade note.
+- **`RunResult.unanswered` → `nonReproducibleAnswers`** — a self-admitted misnomer in the frozen
+  run-result envelope, renamed before the 1.0 freeze.
+- **Body-less manifest entries now carry `ManifestEntry.truncationReason`** (`size` | `readonly` |
+  `unreadable`) instead of a cassette-level `readonlyFolderRoots` list — self-describing and
+  redaction-immune. Replay reads the reason to give `artifact_json` the precise remedy on every lane
+  (assert on a deliverable vs. raise `--max-artifact-bytes`).
+- **`schema/run-result.json` completed** (added `fingerprint`, `userVisibleRoots`, `readonlyFolderRoots`,
+  `preRunPaths`, `toolResults`) and now strict-validated by a test, and the `cassette.v8.json`
+  `ManifestEntry` shape is fully described — so the schemas frozen at 1.0 match what ships. Corrected the
+  git-mode-default documentation (git is the default unless `COWORK_HARNESS_GITSET=0`).
+
 ## [0.25.0] — 2026-07-05
 
 > **Upgrade notes.**

package/README.md

@@ -84,7 +84,7 @@ Only the committed-cassette `replay` above is fully self-contained. Live `run`/`
 > - **Replay only (zero setup):** `cowork-harness replay <cassette>` — no token, no Docker, no agent. The command above.
 > - **`protocol` (real model, no Docker):** needs only the auth token (item 3 below).
 > - **Live `container` / `microvm` / `hostloop` / `cowork`:** needs Docker (or Lima for `microvm`), a staged agent, and the token — run `cowork-harness doctor` first.
-> - **Invocation:** from a source checkout, `node dist/cli.js <cmd>` (or `npm link` to get the `cowork-harness` command); from a global install, `cowork-harness <cmd>`; the companion skill falls back to `npx cowork-harness@>=0.25.0`.
+> - **Invocation:** from a source checkout, `node dist/cli.js <cmd>` (or `npm link` to get the `cowork-harness` command); from a global install, `cowork-harness <cmd>`; the companion skill falls back to `npx cowork-harness@>=0.26.0`.
 >
 > *(One project, three names: the repo folder is `claude-cowork-headless-emulator`; the npm package, the CLI command, and the GitHub repo are all `cowork-harness`.)*
 
@@ -104,7 +104,7 @@ This repo ships a **companion skill** (`.claude/skills/cowork-harness/`) that te
 /plugin install cowork-harness@cowork-harness
 ```
 
-The skill **self-bootstraps the CLI**: if `cowork-harness` isn't on your PATH it falls back to `npx cowork-harness@>=0.25.0` (a version floor that fails loud rather than silently fetching a too-old CLI; Node ≥ 20). Tiers above `protocol` still need Docker/Lima and a Claude Desktop agent binary — see the prerequisites below.
+The skill **self-bootstraps the CLI**: if `cowork-harness` isn't on your PATH it falls back to `npx cowork-harness@>=0.26.0` (a version floor that fails loud rather than silently fetching a too-old CLI; Node ≥ 20). Tiers above `protocol` still need Docker/Lima and a Claude Desktop agent binary — see the prerequisites below.
 
 It also follows the open [Agent Skills](https://github.com/vercel-labs/skills) spec, so it installs cross-editor (Cursor, Codex, OpenCode, …) by pointing the `npx skills` CLI at `.claude/skills/cowork-harness` in this repo. (Working *inside* this repo, the skill auto-loads as a project skill — no install needed.)
 
@@ -119,7 +119,7 @@ A global install is enough for CI `lint`, reading the teaching skill, and replay
 To `run` the worked examples live or copy them as a starting point, use a source checkout. (The marketplace
 skill install itself only pulls `.claude/skills/cowork-harness/` — SKILL.md + `references/` + `scenario.py`/
 assertion keys, per `.claude-plugin/marketplace.json`'s `source` — not the rest of this table; the full set
-above becomes available once the skill's first command self-bootstraps `npx cowork-harness@>=0.25.0` — see
+above becomes available once the skill's first command self-bootstraps `npx cowork-harness@>=0.26.0` — see
 [above](#drive-it-from-claude-code-companion-skill) — which pulls the same npm package as the global-install row.)
 
 ### Prerequisites for anything above `protocol` fidelity

package/SPEC.md

@@ -504,19 +504,25 @@ and a fail observed) always fails that batch, regardless of the numeric rate.
   "egress":    [{ "host","decision":"allow|deny" }],
   "assertions":[{ "assertion": <Assertion>, "pass": bool, "message?": "string" }],
   "subagents": [{ "toolUseId","parentToolUseId?","agentType","declaredTools":[],"toolsUsed":[] }],
-  "unanswered":[{ "question","chosen","by","rationale?","model?" }],
+  "nonReproducibleAnswers?":[{ "question","chosen","by","rationale?","model?" }], // decisions answered by a non-deterministic/non-authoritative source (llm/external/human/first); scripted answers are excluded
   "usage?": { "turns?": number, /* …SDK usage fields (input_tokens, output_tokens, etc.), pass-through */ },
   "cost?": { "usd?": number, "raw?": {...} }, // usd = SDK's total_cost_usd for this invocation; raw = the api_metrics event payload (independent source)
   "durationMs?": number,
+  "fingerprint?": { "baseline", "skillHash?", "skillSources?":[], "skillScope?":[], "sharedHash?", "contentSig?", "fileSigs?":[["relpath","contentSha"]], "fileSigsOmitted?": bool, "mode?": "git|raw", "agentScope?": "skill" }, // skill/plugin staleness fingerprint recorded at run time; lets `verify-run` detect a kept run whose gate snapshot predates a skill change
   "outDir": "string",
   "workDir?": "string",                          // the agent's working root (mnt/) inside the run dir
   "outputsDir?": "string",                       // the user-visible deliverable mount (mnt/outputs)
+  "userVisibleRoots?": ["string"],               // user-visible mount roots (relative to mnt/) — `outputs` plus each connected folder's resolved mount name; plugins excluded
+  "readonlyFolderRoots?": ["string"],            // subset of userVisibleRoots that are read-only (mode:"r") connected-folder mounts — inputs, not deliverables; `artifacts` excludes them
+  "artifacts?": [{ "path","bytes" }],            // files written under the user-visible roots (paths + sizes only — no content snapshot)
+  "preRunPaths?": ["string"],                    // workRoot-relative paths under the user-visible roots that existed BEFORE the agent ran — the `no_unexpected_files` baseline; absent when the tier didn't capture it (microvm) or the run predates the seam
   "effectiveFidelity?": "string",                // tier actually used (differs from `fidelity` when "cowork" resolved)
   "nonDeterministic?": bool,                      // true if any decision came from a non-deterministic source → not reproducible
   "gateProvenance?": { "total": number, "bySource": {…}, "gates": [{ "question","answeredBy","answer","model?" }] }, // how each AskUserQuestion gate was answered; informational (never fails the verdict); live/partial lane only (absent on replay)
   "permissiveAutoAllow?": ["string"],             // tools auto-allowed by cowork parity that real Cowork BLOCKS → green is NOT faithful
   "staleness?": [{ "class": "baseline|skill|shared-root|format|unverifiable-baseline|unverifiable-skill|resolved-tier|unverifiable-tier", "message" }], // replay only; cassette-staleness findings, surfaced for a JSON gate. Non-failing by default (a stale but passing replay stays ok:true); `--strict` fails on every class, `--fail-on-skill-drift` on skill/shared-root/unverifiable-skill only. `resolved-tier` = a `fidelity: cowork` cassette's recorded effectiveFidelity no longer matches the tier the scenario's baseline (pinned `baseline:` or `latest`) resolves to today — the recording exercises the wrong tier; `unverifiable-tier` = the tier check couldn't run for a baseline-dependent (`fidelity: cowork`) cassette (no recorded effectiveFidelity, or its pinned baseline failed to load). Tier resolution is baseline-only (the CLAUDE_FORCE_HOST_LOOP env override is suppressed) so verify results can't differ across machines.
   "skippedAssertions?": { "full": number, "partial": number }, // replay only; count of live-only assertions NOT evaluated (full = whole assertion skipped; partial = content half ran, fs/egress half dropped). The skipped ones are absent from `assertions[]`.
+  "toolResults?": [{ "toolUseId?","isError","text","assertText?" }], // tool-result text at assertion-fidelity cap (10 KB); backs tool_result_contains/tool_result_not_contains
   "skillsInvoked?": ["string"],                  // Wave 1: skill/plugin ids invoked via the Skill tool_use event, call order, duplicates kept. Backs skill_triggered/no_skill_triggered.
   "skillToolAvailable?": bool                     // Wave 1: whether the agent's init tool list included "Skill" — false ⇒ skill_triggered/no_skill_triggered fail as evidence-unavailable (agent-version drift)
 }
@@ -641,9 +647,10 @@ nothing here is guaranteed; minor versions may break any surface — see [RELEAS
   `findings` / `staleness` / `notes` / `version` / `error` channels. This is the machine output the
   CI recipes and the packaged Action steer consumers to parse; renaming or removing a key is
   breaking, adding one is not.
-- **Cassette format** — the current `cassetteVersion` (**7**) and its verdict-modifier assertion keys.
-  Older-version cassettes (the retained `schema/cassette.v2/v3/v5/v6.json` — no v4 schema was ever
-  published) stay replayable; dropping a still-emitted version's readability is breaking.
+- **Cassette format** — the current `cassetteVersion` (**8**, `schema/cassette.v8.json`) and its
+  verdict-modifier assertion keys. Older-version cassettes (the retained
+  `schema/cassette.v2/v3/v5/v6/v7.json` — no v4 schema was ever published) stay replayable; dropping a
+  still-emitted version's readability is breaking.
 - **Control protocol** — `schema/protocol.v1.json` + the golden control-response vectors (§5).
 - **Environment variables** — the documented `COWORK_HARNESS_*` knobs plus `COWORK_AGENT_BINARY` and
   `COWORK_AGENT_IMAGE`. Renaming a documented var or changing its meaning is breaking.

package/dist/assert.js

@@ -166,7 +166,7 @@ function check(a, ctx) {
     const results = [];
     const ok = () => ({ pass: true });
     const fail = (message) => ({ pass: false, message });
-    const truncated = ctx.truncatedPaths ?? new Set();
+    const truncated = ctx.truncatedPaths ?? new Map();
     if (a.transcript_contains !== undefined)
         results.push(ctx.transcriptMissing
             ? fail(`evidence unavailable: transcript sidecar (run.jsonl) absent — cannot evaluate transcript_contains`)
@@ -522,8 +522,14 @@ function check(a, ctx) {
             // where the real file is still on disk. (Existence keys stay green — existence is provable from
             // the recorded hash — but content is genuinely absent, so this fails loud, never vacuous.)
             const rel = relative(resolve(ctx.workRoot), file);
-            const isReadonlyInput = (ctx.readonlyFolderRoots ?? []).some((pre) => rel === pre || rel.startsWith(pre + "/"));
-            const bodyLess = truncated.has(rel) || isReadonlyInput;
+            // Reason sources by lane: LIVE/verify-run derive read-only from the plan's `readonlyFolderRoots`
+            // (no manifest exists at eval time); REPLAY reads the per-entry `truncationReason` off the
+            // materialized manifest (`truncated.get(rel)`). Keeping both is complementary, not redundant.
+            const liveReadonly = (ctx.readonlyFolderRoots ?? []).some((pre) => rel === pre || rel.startsWith(pre + "/"));
+            const replayReason = truncated.get(rel); // undefined if not body-less on replay, or a pre-v8 entry with no reason
+            const isReadonlyInput = liveReadonly || replayReason === "readonly";
+            const isOverCap = replayReason === "size";
+            const bodyLess = truncated.has(rel) || liveReadonly;
             if (!realFile) {
                 results.push(fail(`unsafe artifact_json path "${aj.artifact}" — symlink target escapes the work root`));
             }
@@ -531,14 +537,16 @@ function check(a, ctx) {
                 results.push(fail(`artifact_json: file not found: ${aj.artifact} (under ${ctx.workRoot})`));
             }
             else if (bodyLess) {
-                // Every lane knows WHY the entry is body-less, so the remedy is always precise: live/verify-run
-                // derive `readonlyFolderRoots` from the plan; replay reads the set persisted on the cassette
-                // (v7). A read-only input can't be captured (assert on a deliverable); an over-cap artifact can
-                // (raise the cap). `isReadonlyInput` distinguishes them on all three lanes.
+                // Precise remedy when the cause is known (read-only ⇒ assert on a deliverable; over-cap ⇒ raise
+                // the cap). A pre-v8 entry carries no reason ⇒ name both causes (we can't tell). "unreadable"
+                // also falls here — it's a record-time read failure, so the both-causes text is the safe hint.
+                const cause = isReadonlyInput
+                    ? `(read-only connected-folder input — its content is never captured; assert artifact_json on a deliverable instead)`
+                    : isOverCap
+                        ? `(larger than the artifact-body cap — raise --max-artifact-bytes to capture it)`
+                        : `(a read-only connected-folder input, or an artifact larger than the body cap — if an input, assert on a deliverable; if a large deliverable, raise --max-artifact-bytes)`;
                 results.push(fail(`evidence unavailable: artifact_json target "${aj.artifact}" was captured body-less ` +
-                    (isReadonlyInput
-                        ? `(read-only connected-folder input — its content is never captured; assert artifact_json on a deliverable instead)`
-                        : `(larger than the artifact-body cap — raise --max-artifact-bytes to capture it)`) +
+                    cause +
                     ` — content is not in the cassette, so it cannot be evaluated on replay`));
             }
             else {

package/dist/run/cassette.js

@@ -77,13 +77,14 @@ const MANIFEST_MAX_FILES = 2000;
 // routes pre-v6 cassettes to a re-record (honest "algorithm changed" message, not "content changed").
 // v7: NUL-byte entry separator in hashDir and computeContentSig (replaces `\n`) to prevent hash collisions
 // for file paths containing newline characters.
-export const CASSETTE_VERSION = 7;
+export const CASSETTE_VERSION = 8;
 // The contentSig algorithm version. Bumped whenever computeContentSig's INPUT/encoding changes (the v6
 // unification). `rehash` only byte-compares contentSig within the same algo version; across a bump it
 // re-records. Derived from cassetteVersion: < 6 ⇒ algo 1 (legacy), ≥ 6 ⇒ algo 2 (unified),
-// ≥ 7 ⇒ algo 3 (NUL-byte separator).
-const CONTENTSIG_ALGO = 3;
-const contentSigAlgoOf = (cassetteVersion) => (cassetteVersion >= 7 ? 3 : cassetteVersion >= 6 ? 2 : 1);
+// ≥ 7 ⇒ algo 3 (NUL-byte separator), ≥ 8 ⇒ algo 4 (skillHash folds fixed-length content shas +
+// contentSig/link entries are type-prefixed & NUL-framed — closes unframed-concatenation collisions).
+const CONTENTSIG_ALGO = 4;
+const contentSigAlgoOf = (cassetteVersion) => (cassetteVersion >= 8 ? 4 : cassetteVersion >= 7 ? 3 : cassetteVersion >= 6 ? 2 : 1);
 /** Canonical URL of the JSON Schema for this cassette format version.
  *  Appears in every written cassette as `$schema` so editors and unfamiliar readers
  *  can discover what tool produced the file and what the format means. */
@@ -149,18 +150,23 @@ export function buildManifest(workRoot, cap, roots = ["outputs", ".projects"], b
             abs = containedPath(workRoot, path);
         }
         catch {
-            return { path, bytes, sha256: "", truncated: true };
+            return { path, bytes, sha256: "", truncated: true, truncationReason: "unreadable" };
         }
         let buf;
         try {
             buf = readFileSync(abs);
         }
         catch {
-            return { path, bytes, sha256: "", truncated: true };
+            return { path, bytes, sha256: "", truncated: true, truncationReason: "unreadable" };
         }
         const sha256 = createHash("sha256").update(buf).digest("hex");
-        if (buf.length > limit || isBodyLess(path))
-            return { path, bytes, sha256, truncated: true };
+        // truncationReason names WHY the body is absent so replay can give the precise remedy without a
+        // cassette-level roots list: "readonly" (a mode:r input — assert on a deliverable) vs "size" (over
+        // the body cap — raise --max-artifact-bytes). "unreadable" is the catch branches above.
+        if (isBodyLess(path))
+            return { path, bytes, sha256, truncated: true, truncationReason: "readonly" };
+        if (buf.length > limit)
+            return { path, bytes, sha256, truncated: true, truncationReason: "size" };
         // store an encoding marker. UTF-8-safe bodies stay text (readable cassettes); binary bodies go
         // base64 so the record→replay round-trip is byte-exact and the sha256 verify stays valid.
         if (isLosslessUtf8(buf))
@@ -183,7 +189,9 @@ function decodeBody(e) {
  *  against its recorded sha256 (over the decoded RAW bytes) — a mismatch fails replay (throws). */
 export function materializeManifest(entries, roots = ["outputs", ".projects"]) {
     const workRoot = mkdtempSync(join(tmpdir(), "cwh-replay-"));
-    const truncatedPaths = new Set();
+    // path → why the body is absent (from the entry's truncationReason; `undefined` on a pre-v8 entry that
+    // had no reason). `.has()` still means "is body-less"; `.get()` gives the reason for the precise remedy.
+    const truncatedPaths = new Map();
     for (const e of entries) {
         const abs = containedPath(workRoot, e.path); // reject absolute / `..` / out-of-root before writing
         const raw = decodeBody(e); // decode per the encoding marker
@@ -198,7 +206,7 @@ export function materializeManifest(entries, roots = ["outputs", ".projects"]) {
         mkdirSync(dirname(abs), { recursive: true });
         writeFileSync(abs, raw);
         if (e.truncated)
-            truncatedPaths.add(relative(resolve(workRoot), abs));
+            truncatedPaths.set(relative(resolve(workRoot), abs), e.truncationReason);
     }
     return { workRoot, prefixes: roots, truncatedPaths };
 }
@@ -414,9 +422,6 @@ export function scanCassette(cassette, allow) {
     // cassettes`. Scan them too, prefixed `metadata:` so a reviewer knows redaction here ALSO rewrites
     // structural paths, distinct from free-text findings in the transcript/deliverable.
     (cassette.userVisibleRoots ?? []).forEach((r, i) => findings.push(...scanText(r, `metadata:userVisibleRoots[${i}]`, allow, FULL)));
-    // readonlyFolderRoots is a subset of userVisibleRoots (same strings), but it's a separately-persisted
-    // field — scan it too so a reviewer can't be surprised by an unscanned persisted string.
-    (cassette.readonlyFolderRoots ?? []).forEach((r, i) => findings.push(...scanText(r, `metadata:readonlyFolderRoots[${i}]`, allow, FULL)));
     findings.push(...scanText(cassette.scenario.name ?? "", "metadata:scenario.name", allow, FULL));
     findings.push(...scanText(cassette.scenario.session ?? "", "metadata:scenario.session", allow, FULL));
     if (cassette.scenarioSource)
@@ -1024,15 +1029,10 @@ export function redactCassette(cassette, policy) {
         }
     }
     const redactedPreRunPaths = cassette.preRunPaths?.map((p) => redactText(p, policy));
-    // readonlyFolderRoots is a subset of userVisibleRoots; redact it with the same context-free redactText
-    // so the prefix relationship with the (redacted) artifact paths is preserved. The `...cassette` spread
-    // would otherwise carry the UNREDACTED set through.
-    const redactedReadonly = cassette.readonlyFolderRoots?.map((r) => redactText(r, policy));
     return {
         ...cassette,
         scenario,
         userVisibleRoots: redactedRoots,
-        readonlyFolderRoots: redactedReadonly,
         artifacts: redactedArtifacts,
         preRunPaths: redactedPreRunPaths,
         events: cassette.events.map((l) => redactJsonLine(l, policy)),
@@ -1321,11 +1321,15 @@ export function redactionPreflightMessage(items) {
  *  verify-run, and replay (evidence-unavailable, see assert.ts) — so they are neither an asymmetry nor a
  *  "too large" problem. Paths are normalized through `resolve` so `./outputs/x.json` and `outputs/x.json`
  *  join cleanly against the manifest's walk paths. */
-export function artifactJsonTargetsTruncated(scenario, workRoot, artifacts, readonlyFolderRoots = []) {
-    const isReadonly = (p) => readonlyFolderRoots.some((pre) => p === pre || p.startsWith(pre + "/"));
+export function artifactJsonTargetsTruncated(scenario, workRoot, artifacts) {
+    // Flag ONLY size-truncated entries (`truncationReason === "size"`) — the genuine green-record/red-replay
+    // case whose remedy is "raise the cap". "readonly"/"unreadable" are excluded: raising the cap can't
+    // capture them, and artifact_json against one already fails loud-and-symmetrically (evidence-unavailable,
+    // assert.ts). (A pre-v8 entry with no reason is not flagged — this guard only runs at record time, where
+    // buildManifest always sets the reason.)
     const truncatedAbs = new Set();
     for (const a of artifacts)
-        if (a.truncated && !isReadonly(a.path))
+        if (a.truncated && a.truncationReason === "size")
             truncatedAbs.add(resolve(workRoot, a.path));
     if (truncatedAbs.size === 0)
         return [];
@@ -1868,7 +1872,7 @@ async function recordScenarioObject(scenario, opts, extraPolicyDirs = []) {
     // replay (no committed body). Surface that record→replay asymmetry NOW, at its cause, instead of letting a
     // green record produce a red replay in CI. Honor --allow-failing (warn, don't block) like the verdict gate.
     if (result.workDir) {
-        const truncatedAsserted = artifactJsonTargetsTruncated(scenario, result.workDir, artifacts, result.readonlyFolderRoots ?? []);
+        const truncatedAsserted = artifactJsonTargetsTruncated(scenario, result.workDir, artifacts);
         if (truncatedAsserted.length) {
             const cap = opts.maxArtifactBytes ?? defaultBodyCap();
             const msg = `artifact_json asserts artifact(s) too large to commit (>${cap} B, stored hash-only): ${truncatedAsserted.join(", ")} — ` +
@@ -1890,11 +1894,10 @@ async function recordScenarioObject(scenario, opts, extraPolicyDirs = []) {
         effectiveFidelity: result.effectiveFidelity,
         artifacts,
         userVisibleRoots: recordRoots,
-        // Persist the read-only subset ONLY when non-empty (keeps cassettes without connected folders
-        // clean); replay reads it to distinguish a body-less read-only input from an over-cap artifact.
-        ...(result.readonlyFolderRoots?.length ? { readonlyFolderRoots: result.readonlyFolderRoots } : {}),
-        // co-present with userVisibleRoots by construction (see the Cassette field comment) — a cassette
-        // carrying preRunPaths never hits replay's legacy-roots fallback.
+        // (v8: no cassette-level readonlyFolderRoots — the read-only reason rides per-entry on
+        // ManifestEntry.truncationReason, set by buildManifest above from RunResult.readonlyFolderRoots.)
+        // co-present with userVisibleRoots by construction — a cassette carrying preRunPaths never hits
+        // replay's legacy-roots fallback.
         preRunPaths: result.preRunPaths,
         // persist the authored scenario source file RELATIVE to the cassette dir (relocatable, no
         // absolute host path) so `--rerecord-stale` re-records from the edited YAML even when name ≠ filename.
@@ -2713,7 +2716,7 @@ export async function replayCassette(cassette, hooks = [], opts = {}) {
     }
     const { workRoot: replayWorkRoot, prefixes: replayPrefixes, truncatedPaths: replayTruncatedPaths, } = manifestKeys.length
         ? materializeManifest(cassette.artifacts, cassette.userVisibleRoots ?? ["outputs", ".projects"])
-        : { workRoot: "", prefixes: [], truncatedPaths: new Set() };
+        : { workRoot: "", prefixes: [], truncatedPaths: new Map() };
     // materializeManifest created a temp dir (`replayWorkRoot`) above; everything below uses it and
     // then returns. Wrap the rest in try/finally so the temp dir is removed on every exit path (normal
     // return OR a throw from evaluate/assert building) — otherwise `cwh-replay-*` dirs leak under tmpdir
@@ -2780,9 +2783,10 @@ export async function replayCassette(cassette, hooks = [], opts = {}) {
             result: rec.result,
             workRoot: replayWorkRoot,
             userVisiblePrefixes: replayPrefixes,
-            // Persisted on the cassette (v7 optional metadata) so replay tells a body-less read-only input
-            // apart from an over-cap artifact and gives the precise artifact_json remediation.
-            readonlyFolderRoots: cassette.readonlyFolderRoots ?? [],
+            // Replay reads the body-less REASON per-entry from the materialized manifest (truncatedPaths, a
+            // Map<path, reason>) — NOT a cassette-level roots list (removed in v8). Live/verify-run alone use
+            // readonlyFolderRoots (they have no manifest at eval time), so it's empty here.
+            readonlyFolderRoots: [],
             preRunPaths: cassette.preRunPaths,
             outputsDeletes: [],
             questions: rec.questions,
@@ -2896,7 +2900,7 @@ export async function replayCassette(cassette, hooks = [], opts = {}) {
             egress: [],
             assertions,
             subagents: rec.subagents,
-            unanswered: rec.unanswered,
+            nonReproducibleAnswers: rec.unanswered,
             // Wave 0 seam: the live/success/partial assemblers already passed these through; replay never did,
             // so a cassette that recorded usage/cost silently dropped it on replay. re-drive (rec) recomputes
             // them deterministically from the same events, so this is a content key, not a live-only one.

package/dist/run/execute.js

@@ -780,7 +780,7 @@ export async function executeScenario(scenario, opts = {}) {
         assertions,
         toolResults: record.toolResults,
         subagents: record.subagents,
-        unanswered: record.unanswered,
+        nonReproducibleAnswers: record.unanswered,
         usage: record.usage,
         cost: record.cost,
         skillsInvoked: record.skillsInvoked,
@@ -948,7 +948,7 @@ export function buildPartialResult(args) {
         assertions: [],
         toolResults: record.toolResults,
         subagents: record.subagents,
-        unanswered: record.unanswered,
+        nonReproducibleAnswers: record.unanswered,
         usage: record.usage,
         cost: record.cost,
         skillsInvoked: record.skillsInvoked,

package/dist/run/renderer.js

@@ -282,7 +282,7 @@ function renderGateProvenance(r, plan, write) {
         write(`   ${dim(plan, line)}\n`);
 }
 function renderAnswerHints(r, plan, write) {
-    const u = r.unanswered ?? [];
+    const u = r.nonReproducibleAnswers ?? [];
     if (!u.length)
         return;
     // partition by the REAL `by:` vocabulary (decider.ts:25 — scripted|cowork|strict|human|agent|

package/dist/run/skill-hash.js

@@ -125,7 +125,8 @@ function hashDir(dir, hash, errors, rel = "", accept, onFile, root = dir) {
                 continue;
             }
             const targetRel = relative(rootResolved, resolvedTarget).split(sep).join("/");
-            hash.update(`L:${relPath} -> ${targetRel}\0`); // link structure; a re-point changes the digest
+            hash.update(`L:${relPath}\0${targetRel}\0`); // link structure; a re-point changes the digest. NUL-separate
+            // path from target (not ` -> `): otherwise `a`->`b -> c` and `a -> b`->`c` both fold `L:a -> b -> c\0`.
             if (onFile)
                 onFile(relPath, `lnk:${targetRel}`);
             continue;
@@ -181,9 +182,15 @@ function hashDir(dir, hash, errors, rel = "", accept, onFile, root = dir) {
                     /* not valid JSON — fall through to raw-byte hashing */
                 }
             }
-            hash.update(hashedContent);
+            // Fold the fixed-length content SHA, NOT the raw content: raw bytes have no length/terminator, so
+            // `F:a\0<A>F:b\0<B>` would collide with a single file `a` whose content is `<A>F:b\0<B>` (a staleness
+            // false-negative). A 64-hex sha is self-delimiting and its charset is disjoint from the `F:`/`L:`/`D:`
+            // prefixes, closing the unframed-concatenation hole. Hoisted out of `if (onFile)` so the digest is
+            // always folded (the no-sink path must fold the same bytes).
+            const contentSha = createHash("sha256").update(hashedContent).digest("hex");
+            hash.update(contentSha);
             if (onFile)
-                onFile(relPath, createHash("sha256").update(hashedContent).digest("hex"));
+                onFile(relPath, contentSha);
         }
     }
 }
@@ -295,7 +302,10 @@ function scopedAccept(keep, dirSkills, scopeAgents) {
 export function computeContentSig(dirs, scopeSkills, sessionIgnore) {
     if (dirs.length === 0)
         return undefined;
-    const entries = skillHashEntries(dirs, scopeSkills, sessionIgnore).map((e) => `${e.path}:${e.sha}`);
+    // Type-prefix + NUL-separate path from sha: without them, a file named `a:lnk` with content-sha
+    // `<64hex>` collides with a symlink `a` -> target `<64hex>` (both `a:lnk:<64hex>`). `e.sha` starts with
+    // `lnk:` only for links (a 64-hex file sha cannot — l/n/k/: are non-hex), so it disambiguates the kind.
+    const entries = skillHashEntries(dirs, scopeSkills, sessionIgnore).map((e) => e.sha.startsWith("lnk:") ? `L:${e.path}\0${e.sha}` : `F:${e.path}\0${e.sha}`);
     if (entries.length === 0)
         return undefined;
     return createHash("sha256").update(entries.join("\0")).digest("hex");

package/docs/cassette.md

@@ -48,18 +48,17 @@ re-record). `expect_denied`/filesystem/egress keys are sourced but stay live-onl
 
 ```jsonc
 {
-  "cassetteVersion": 7,                  // format version; ABSENT = legacy (0); a FUTURE version warns
+  "cassetteVersion": 8,                  // format version; ABSENT = legacy (0); a FUTURE version warns
   "scenario": { /* Scenario object — same schema as the .yaml */ },
   "events": [ /* JSON lines from events.jsonl (child→driver stdout) */ ],
   "controlOut": [ /* JSON lines from control-out.jsonl (driver→child control_responses) */ ],
   "userVisibleRoots": ["outputs", "myproject"], // visible roots = outputs + each connected folder's mount name (its basename; `.projects` is the pre-1.14271.0 legacy fallback)
-  "readonlyFolderRoots": ["carta-folder"], // the mode:r subset of userVisibleRoots (body-less); lets replay give artifact_json the precise remedy. Optional; absent when no read-only folder
   "artifacts": [                         // snapshot of outputs/ + connected folders (optional)
     { "path": "outputs/x.json", "bytes": 24, "sha256": "…", "body": "{…}" }, // body inlined ≤ 64 KiB
-    { "path": "outputs/big.bin", "bytes": 9e6, "sha256": "…", "truncated": true }, // oversized → hash-only
-    { "path": "carta-folder/input.xlsx", "bytes": 4096, "sha256": "…", "truncated": true } // mode:r connected-folder INPUT → body-less (see below), regardless of size
+    { "path": "outputs/big.bin", "bytes": 9e6, "sha256": "…", "truncated": true, "truncationReason": "size" }, // oversized → hash-only (raise --max-artifact-bytes)
+    { "path": "carta-folder/input.xlsx", "bytes": 4096, "sha256": "…", "truncated": true, "truncationReason": "readonly" } // mode:r connected-folder INPUT → body-less (see below), regardless of size
   ],
-  "fingerprint": { "baseline": "1.15962.1", "skillHash": "…", "mode": "git", "contentSig": "…", "fileSigs": [["skills/x/SKILL.md", "…"]], "skillSources": ["…"] }, // staleness tripwire (v5: fileSigs only; v6: mode + git default; v7: NUL-delimited hash entries)
+  "fingerprint": { "baseline": "1.15962.1", "skillHash": "…", "mode": "git", "contentSig": "…", "fileSigs": [["skills/x/SKILL.md", "…"]], "skillSources": ["…"] }, // staleness tripwire (v5: fileSigs only; v6: mode + git default; v7: NUL-delimited hash entries; v8: folds fixed-length content shas + type-prefixed/NUL-framed entries)
   "authoring": { "nonDeterministic": true, "channel": "decider-dir" } // present ONLY when a live decider answered ≥1 gate (see §Answering gates during recording); re-record may drift, replay is still deterministic
 }
 ```
@@ -390,7 +389,10 @@ possible.
 - **`recorded under an older hash format (vN → vM)`** — format upgrade; re-record once and the message
   goes away. (Cassettes recorded before format **v6** all need one re-record — see the boundary note below.
   Cassettes recorded at **v6** need one re-record after upgrading to the v7 format, which switched the
-  hash-entry delimiter from `\n` to `\0`.)
+  hash-entry delimiter from `\n` to `\0`. Cassettes recorded at **v7** need one re-record after upgrading
+  to **v8**, which folds fixed-length content shas and type-prefixes/NUL-frames the hash entries — a v7
+  fingerprint is non-comparable, so `rehash` routes v7 cassettes to re-record. A cassette that carries no
+  `skillHash` is unaffected and keeps replaying.)
 - **`skill files changed since record — N changed (path, …)`** — the **exact** changed/added/removed file(s),
   from the per-file manifest (`fileSigs`). For a scoped cassette the drift is attributed **per bucket** by the
   actual changed paths: a `shared root changed (scope: skills/x) [N changed (…)]` message for shared-dependency

package/examples/replays/README.md

@@ -13,7 +13,7 @@ replay path end-to-end, including:
 
 Run it with:
 
-> Assumes the `cowork-harness` CLI is available — from a source checkout run `npm ci && npm run build && npm link` first, or `npm i -g cowork-harness@>=0.25.0`. (`replay` itself needs nothing else — no token, no Docker.)
+> Assumes the `cowork-harness` CLI is available — from a source checkout run `npm ci && npm run build && npm link` first, or `npm i -g cowork-harness@>=0.26.0`. (`replay` itself needs nothing else — no token, no Docker.)
 
 ```sh
 cowork-harness replay examples/replays/example-pdf-skill.cassette.json

package/examples/replays/example-pdf-skill.cassette.json

@@ -1,7 +1,7 @@
 {
-  "$schema": "https://raw.githubusercontent.com/yaniv-golan/cowork-harness/main/schema/cassette.v7.json",
+  "$schema": "https://raw.githubusercontent.com/yaniv-golan/cowork-harness/main/schema/cassette.v8.json",
   "generator": "cowork-harness",
-  "cassetteVersion": 7,
+  "cassetteVersion": 8,
   "scenario": {
     "name": "example-pdf-skill",
     "baseline": "latest",
@@ -46,28 +46,31 @@
   },
   "events": [
     "{\"type\":\"control_response\",\"response\":{\"subtype\":\"success\",\"request_id\":\"init-1\",\"response\":{\"commands\":[{\"name\":\"deep-research\",\"description\":\"Deep research harness — fan-out web searches, fetch sources, adversarially verify claims, synthesize a cited report. (dynamic workflow)\",\"argumentHint\":\"\"},{\"name\":\"my-pdf-skill\",\"description\":\"(my-pdf-skill) Example skill under test. Summarize a PDF and write action items.\",\"argumentHint\":\"\"},{\"name\":\"clear\",\"description\":\"Start a new session with empty context; previous session stays on disk (resumable with /resume)\",\"argumentHint\":\"[name]\",\"aliases\":[\"reset\",\"new\"]},{\"name\":\"compact\",\"description\":\"Free up context by summarizing the conversation so far\",\"argumentHint\":\"<optional custom summarization instructions>\"},{\"name\":\"config\",\"description\":\"Set a setting by key\",\"argumentHint\":\"key=value\",\"aliases\":[\"settings\"]},{\"name\":\"context\",\"description\":\"Show current context usage\",\"argumentHint\":\"\"},{\"name\":\"heapdump\",\"description\":\"Dump the JS heap to ~/Desktop\",\"argumentHint\":\"\"},{\"name\":\"init\",\"description\":\"Initialize a new CLAUDE.md file with codebase documentation\",\"argumentHint\":\"\"},{\"name\":\"reload-skills\",\"description\":\"Pick up skills added or changed on disk during this session\",\"argumentHint\":\"\"},{\"name\":\"review\",\"description\":\"Review a GitHub pull request; for your working diff use /code-review\",\"argumentHint\":\"[pr number]\"},{\"name\":\"security-review\",\"description\":\"Complete a security review of the pending changes on the current branch\",\"argumentHint\":\"\"},{\"name\":\"usage\",\"description\":\"Show session cost, plan usage, and what's contributing to your limits\",\"argumentHint\":\"\",\"aliases\":[\"cost\",\"stats\"]},{\"name\":\"insights\",\"description\":\"Generate a report analyzing your Claude Code sessions\",\"argumentHint\":\"\"},{\"name\":\"goal\",\"description\":\"Set a goal — keep working until the condition is met\",\"argumentHint\":\"\"},{\"name\":\"team-onboarding\",\"description\":\"Help teammates ramp on Claude Code with a guide from your usage\",\"argumentHint\":\"\"}],\"agents\":[{\"name\":\"claude\",\"description\":\"Catch-all for any task that doesn't fit a more specific agent. FleetView's default when no agent name is typed.\"},{\"name\":\"claude-code-guide\",\"description\":\"Use this agent when the user asks questions (\\\"Can Claude...\\\", \\\"Does Claude...\\\", \\\"How do I...\\\") about: (1) Claude Code (the CLI tool) - features, hooks, slash commands, MCP servers, settings, IDE integrations, keyboard shortcuts; (2) Claude Agent SDK - building custom agents; (3) Claude API (formerly Anthropic API) - API usage, tool use, Anthropic SDK usage. **IMPORTANT:** Before spawning a new agent, check if there is already a running or recently completed claude-code-guide agent that you can continue via SendMessage.\",\"model\":\"haiku\"},{\"name\":\"Explore\",\"description\":\"Fast read-only search agent for locating code. Use it to find files by pattern (eg. \\\"src/components/**/*.tsx\\\"), grep for symbols or keywords (eg. \\\"API endpoints\\\"), or answer \\\"where is X defined / which files reference Y.\\\" Do NOT use it for code review, design-doc auditing, cross-file consistency checks, or open-ended analysis — it reads excerpts rather than whole files and will miss content past its read window. When calling, specify search breadth: \\\"quick\\\" for a single targeted lookup, \\\"medium\\\" for moderate exploration, or \\\"very thorough\\\" to search across multiple locations and naming conventions.\",\"model\":\"haiku\"},{\"name\":\"general-purpose\",\"description\":\"General-purpose agent for researching complex questions, searching for code, and executing multi-step tasks. When you are searching for a keyword or file and are not confident that you will find the right match in the first few tries use this agent to perform the search for you.\"},{\"name\":\"Plan\",\"description\":\"Software architect agent for designing implementation plans. Use this when you need to plan the implementation strategy for a task. Returns step-by-step plans, identifies critical files, and considers architectural trade-offs.\"},{\"name\":\"statusline-setup\",\"description\":\"Use this agent to configure the user's Claude Code status line setting.\",\"model\":\"sonnet\"}],\"output_style\":\"default\",\"available_output_styles\":[\"default\",\"Proactive\",\"Explanatory\",\"Learning\"],\"models\":[{\"value\":\"default\",\"resolvedModel\":\"claude-sonnet-5\",\"displayName\":\"Default (recommended)\",\"description\":\"Sonnet 5 · Efficient for routine tasks\",\"supportsEffort\":true,\"supportedEffortLevels\":[\"low\",\"medium\",\"high\",\"xhigh\",\"max\"],\"supportsAdaptiveThinking\":true,\"supportsAutoMode\":true},{\"value\":\"sonnet\",\"resolvedModel\":\"claude-sonnet-5\",\"displayName\":\"Sonnet\",\"description\":\"Sonnet 5 · Efficient for routine tasks\",\"supportsEffort\":true,\"supportedEffortLevels\":[\"low\",\"medium\",\"high\",\"xhigh\",\"max\"],\"supportsAdaptiveThinking\":true,\"supportsAutoMode\":true},{\"value\":\"opus\",\"resolvedModel\":\"claude-opus-4-8\",\"displayName\":\"Opus\",\"description\":\"Opus 4.8 · Best for everyday, complex tasks\",\"supportsEffort\":true,\"supportedEffortLevels\":[\"low\",\"medium\",\"high\",\"xhigh\",\"max\"],\"supportsAdaptiveThinking\":true,\"supportsFastMode\":true,\"supportsAutoMode\":true},{\"value\":\"haiku\",\"resolvedModel\":\"claude-haiku-4-5-20251001\",\"displayName\":\"Haiku\",\"description\":\"Haiku 4.5 · Fastest for quick answers\"}],\"account\":{\"tokenSource\":\"CLAUDE_CODE_OAUTH_TOKEN\",\"apiProvider\":\"firstParty\"},\"pid\":1}}}",
-    "{\"type\":\"system\",\"subtype\":\"init\",\"cwd\":\"/sessions/local_9hldogcxp\",\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"tools\":[\"Task\",\"AskUserQuestion\",\"Bash\",\"Edit\",\"Glob\",\"Grep\",\"NotebookEdit\",\"Read\",\"Skill\",\"TaskCreate\",\"TaskGet\",\"TaskList\",\"TaskStop\",\"TaskUpdate\",\"ToolSearch\",\"WebFetch\",\"WebSearch\",\"Write\"],\"mcp_servers\":[],\"model\":\"claude-opus-4-8\",\"permissionMode\":\"default\",\"slash_commands\":[\"deep-research\",\"my-pdf-skill:my-pdf-skill\",\"clear\",\"compact\",\"config\",\"context\",\"heapdump\",\"init\",\"reload-skills\",\"review\",\"security-review\",\"usage\",\"insights\",\"goal\",\"team-onboarding\"],\"apiKeySource\":\"none\",\"claude_code_version\":\"2.1.197\",\"output_style\":\"default\",\"agents\":[\"claude\",\"claude-code-guide\",\"Explore\",\"general-purpose\",\"Plan\",\"statusline-setup\"],\"skills\":[\"deep-research\",\"my-pdf-skill:my-pdf-skill\"],\"plugins\":[{\"name\":\"my-pdf-skill\",\"path\":\"/sessions/local_9hldogcxp/mnt/.local-plugins/marketplaces/local-desktop-app-uploads/my-pdf-skill\",\"source\":\"my-pdf-skill@inline\"}],\"analytics_disabled\":false,\"product_feedback_disabled\":false,\"uuid\":\"b92dc0fe-2f0e-4cfc-ae0f-e1cef20650f3\",\"memory_paths\":{\"auto\":\"/sessions/local_9hldogcxp/mnt/.claude/projects/-sessions-local-9hldogcxp/memory/\"},\"fast_mode_state\":\"off\"}",
-    "{\"type\":\"system\",\"subtype\":\"thinking_tokens\",\"estimated_tokens\":50,\"estimated_tokens_delta\":50,\"uuid\":\"7e52a15d-a4d1-488d-8349-494cbea01573\",\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\"}",
-    "{\"type\":\"system\",\"subtype\":\"thinking_tokens\",\"estimated_tokens\":93,\"estimated_tokens_delta\":43,\"uuid\":\"674661a8-1e6a-4c07-a1b0-4c3f6eeaeca6\",\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\"}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_011NpTDBTKfu6acrgobjx7wJ\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"thinking\",\"thinking\":\"\",\"signature\":\"EuwCCmMIDxgCKkBhvNM1ibwEssNe5nhyywDi5SMLX1TvVV1sXRbB7+WfFv7dety7U7IhPLoqvE3ijcoNuOgpAq9ZXxb64pOMfoUTMg9jbGF1ZGUtb3B1cy00LTg4AEIIdGhpbmtpbmcSDFekfDPrlzJoz/a5ORoMsnPUPG/c1lh1XvbUIjAhoW4EbMGsQ7BSI3fQI8AV0n+8Ongs6OPdIH07WnmSXVRAdwB736SxeDf8aAYqLqoqtgG2NHSAHtbFflkdU7UG1CIWpHWww8iXBfJABxUShX6m31EMD53oiJERttchU0fE3Ybo/lwsCMiCAMmcaMIjFuiYz9bg6UpPXc74X0LPjTfxmB9cIG62CCzHgPH1+c6+kebPCC2xkrP8V2Js9JIuSSLCqsguK8Pj+ZZ7vG9jeRwTIoRwQzc/in2aqRwVAYPyFpJxe7oRJs7lhPFBHojp1e7cchCMuiOVPrAKhx7eOAaRNe5zlCXCUhgB\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":10994,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":10994},\"output_tokens\":5,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"1907d44d-9279-4d5a-96e8-188d275b8460\",\"request_id\":\"req_011Cce4gV8iD2cRhnK9HXmvZ\"}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_011NpTDBTKfu6acrgobjx7wJ\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"I'll use the relevant skill for this task. Let me start by reading it.\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":10994,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":10994},\"output_tokens\":5,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"4115500d-7909-47ad-b6bb-8fd1202b9818\",\"request_id\":\"req_011Cce4gV8iD2cRhnK9HXmvZ\"}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_011NpTDBTKfu6acrgobjx7wJ\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01W1bQt5QwmpJV729hM5Gx2n\",\"name\":\"Skill\",\"input\":{\"skill\":\"my-pdf-skill:my-pdf-skill\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":10994,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":10994},\"output_tokens\":5,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"725fc121-8198-4a03-a60f-c4dde3018938\",\"request_id\":\"req_011Cce4gV8iD2cRhnK9HXmvZ\"}",
-    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"tool_result\",\"tool_use_id\":\"toolu_01W1bQt5QwmpJV729hM5Gx2n\",\"content\":\"Launching skill: my-pdf-skill:my-pdf-skill\"}]},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"1150d45d-ddb7-49af-ae40-3cc761cde075\",\"timestamp\":\"2026-07-02T23:02:20.283Z\",\"tool_use_result\":{\"success\":true,\"commandName\":\"my-pdf-skill:my-pdf-skill\"}}",
-    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"text\",\"text\":\"Base directory for this skill: /sessions/local_9hldogcxp/mnt/.local-plugins/marketplaces/local-desktop-app-uploads/my-pdf-skill/skills/my-pdf-skill\\n\\n# my-pdf-skill\\nPlaceholder skill so the example scenario resolves. Replace with the skill you're testing.\\n\"}]},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"4c90d132-e985-4e74-bcba-f54b7d37550f\",\"timestamp\":\"2026-07-02T23:02:20.281Z\",\"isSynthetic\":true}",
-    "{\"type\":\"rate_limit_event\",\"rate_limit_info\":{\"status\":\"allowed\",\"resetsAt\":1783039800,\"rateLimitType\":\"five_hour\",\"overageStatus\":\"allowed\",\"overageResetsAt\":1783033200,\"isUsingOverage\":false},\"uuid\":\"238a9fae-804b-49c9-9f0a-36fe13470668\",\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\"}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_014joZFXLAU5nNqqD7kjt25f\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"thinking\",\"thinking\":\"\",\"signature\":\"EusCCmMIDxgCKkA8IK1poQWDbzDhcKw+Cd9wDhCB9QhBhezUtvRwrauiEknnJmHcEkEnANBsCNwmxA1bn5MckHzGdnKORqHnx19mMg9jbGF1ZGUtb3B1cy00LTg4AEIIdGhpbmtpbmcSDHlxpJop9LVkHZpRMhoMyeT50lP8zV59OSKnIjChDwFGGfvHsWmIvXdjILszWVmfHR27P8W+BwZamrMek8w2n/5AjfwlDXeEC8PkQY8qtQHCIbxmjW45owFmDvmuP8I34n1ywdR8ZlRgX19zUx1zVhUjwZ5+b0Qswzw+4suegTqgAaEF3V33KkrB0DzBACbUH0OvgxfnKSxyryKWJJNfEe/zSeCNLjnPTXgjORf+V5LPMB+ZUw9hjrQCe9MLycYbxJlLqbU9QIPEWWF2G1AT901dLG7EkTqikQoOYggpWpUlguFYhU7H4ZqPOaHM1ilNr50ClZ6L1ObrEUkjz4MblQ2frnVIGAE=\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1,\"cache_creation_input_tokens\":1481,\"cache_read_input_tokens\":10994,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1481},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"a7afbae0-669e-480f-96f7-be2c9eefe508\",\"request_id\":\"req_011Cce4ghR9NBMRfH96ZK6Fd\"}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_014joZFXLAU5nNqqD7kjt25f\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01GT8S8cVVeubi74wfsKzZBR\",\"name\":\"Bash\",\"input\":{\"command\":\"ls -la /sessions/local_9hldogcxp/mnt/uploads/\",\"description\":\"List uploaded files\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1,\"cache_creation_input_tokens\":1481,\"cache_read_input_tokens\":10994,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1481},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"4b2adc91-61ac-48ee-b082-222fce657c04\",\"request_id\":\"req_011Cce4ghR9NBMRfH96ZK6Fd\"}",
-    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01GT8S8cVVeubi74wfsKzZBR\",\"type\":\"tool_result\",\"content\":\"total 4\\ndrwxr-xr-x 3 ubuntu ubuntu   96 Jul  2 23:02 .\\ndrwxr-xr-x 9 ubuntu ubuntu  288 Jul  2 23:02 ..\\n-rw-r--r-- 1 ubuntu ubuntu 1056 Jul  2 23:02 report.pdf\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"60e1eed5-40ba-4183-ba69-8d1c583111b8\",\"timestamp\":\"2026-07-02T23:02:23.509Z\",\"tool_use_result\":{\"stdout\":\"total 4\\ndrwxr-xr-x 3 ubuntu ubuntu   96 Jul  2 23:02 .\\ndrwxr-xr-x 9 ubuntu ubuntu  288 Jul  2 23:02 ..\\n-rw-r--r-- 1 ubuntu ubuntu 1056 Jul  2 23:02 report.pdf\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false}}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01KfNUCMF6CJct9VmhNDaD98\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_016LMRfWpVHmtqBJ2t9XUhp4\",\"name\":\"Read\",\"input\":{\"file_path\":\"/sessions/local_9hldogcxp/mnt/uploads/report.pdf\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":280,\"cache_read_input_tokens\":12475,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":280},\"output_tokens\":72,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"713e27b2-f030-4570-8065-dfc9e4b1f003\",\"request_id\":\"req_011Cce4gw9PXKuEiWx5aABXh\"}",
-    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_016LMRfWpVHmtqBJ2t9XUhp4\",\"type\":\"tool_result\",\"content\":\"PDF file read: /sessions/local_9hldogcxp/mnt/uploads/report.pdf (1KB)\"}]},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"1afafeec-7e0f-4dfb-a4f2-e962ab15ef2f\",\"timestamp\":\"2026-07-02T23:02:25.297Z\",\"tool_use_result\":{\"type\":\"pdf\",\"file\":{\"filePath\":\"/sessions/local_9hldogcxp/mnt/uploads/report.pdf\",\"base64\":\"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\",\"originalSize\":1056}}}",
-    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"document\",\"source\":{\"type\":\"base64\",\"media_type\":\"application/pdf\",\"data\":\"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\"}}]},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"84b68b76-82ad-4c5f-9045-0650738593b7\",\"timestamp\":\"2026-07-02T23:02:25.296Z\",\"isSynthetic\":true}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01Sg8FoDGdjfjS7c344G8LNq\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01TtaFiXBUMRG5YL7NXeD5nA\",\"name\":\"Write\",\"input\":{\"file_path\":\"/sessions/local_9hldogcxp/mnt/project/outputs/actions.md\",\"content\":\"# Q2 Project Status Report — Action Items\\n\\n- [ ] **Alice:** Finalize the production rollout plan by June 20.\\n- [ ] **Bob:** Add monitoring dashboards for the new pipeline.\\n- [ ] **Carol:** Update the on-call runbook and notify support.\\n\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":1844,\"cache_read_input_tokens\":12755,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1844},\"output_tokens\":39,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"44365f11-bcfa-470c-8323-35a2bdc82e87\",\"request_id\":\"req_011Cce4h4o1yerWEFsmfudQN\"}",
-    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01TtaFiXBUMRG5YL7NXeD5nA\",\"type\":\"tool_result\",\"content\":\"File created successfully at: /sessions/local_9hldogcxp/mnt/project/outputs/actions.md (file state is current in your context — no need to Read it back)\"}]},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"cfd9f4b0-78ec-4d0b-bebc-73b041a1387d\",\"timestamp\":\"2026-07-02T23:02:28.905Z\",\"tool_use_result\":{\"type\":\"create\",\"filePath\":\"/sessions/local_9hldogcxp/mnt/project/outputs/actions.md\",\"content\":\"# Q2 Project Status Report — Action Items\\n\\n- [ ] **Alice:** Finalize the production rollout plan by June 20.\\n- [ ] **Bob:** Add monitoring dashboards for the new pipeline.\\n- [ ] **Carol:** Update the on-call runbook and notify support.\\n\",\"structuredPatch\":[],\"originalFile\":null,\"userModified\":false}}",
-    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01Xc6QxbKMXK477dkNP4jffj\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"Done. Here's a summary of **report.pdf** (Q2 Project Status Report):\\n\\nThe migration to the new ingest pipeline is **on track**. Key results:\\n- End-to-end latency dropped **30%** after the caching layer landed.\\n- Error rates are **within target**.\\n- Remaining risk: the **data-backfill window**.\\n\\nI wrote the three action items to `outputs/actions.md` as a markdown checklist:\\n\\n- [ ] **Alice:** Finalize the production rollout plan by June 20.\\n- [ ] **Bob:** Add monitoring dashboards for the new pipeline.\\n- [ ] **Carol:** Update the on-call runbook and notify support.\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":255,\"cache_read_input_tokens\":14599,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":255},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"uuid\":\"20610d4e-eff3-44a4-a822-7244cb449e3e\",\"request_id\":\"req_011Cce4hLHvvA7rGvcWNJPKj\"}",
-    "{\"type\":\"result\",\"subtype\":\"success\",\"is_error\":false,\"api_error_status\":null,\"duration_ms\":15180,\"duration_api_ms\":15069,\"ttft_ms\":2600,\"ttft_stream_ms\":1357,\"time_to_request_ms\":36,\"num_turns\":7,\"result\":\"Done. Here's a summary of **report.pdf** (Q2 Project Status Report):\\n\\nThe migration to the new ingest pipeline is **on track**. Key results:\\n- End-to-end latency dropped **30%** after the caching layer landed.\\n- Error rates are **within target**.\\n- Remaining risk: the **data-backfill window**.\\n\\nI wrote the three action items to `outputs/actions.md` as a markdown checklist:\\n\\n- [ ] **Alice:** Finalize the production rollout plan by June 20.\\n- [ ] **Bob:** Add monitoring dashboards for the new pipeline.\\n- [ ] **Carol:** Update the on-call runbook and notify support.\",\"stop_reason\":\"end_turn\",\"session_id\":\"35ccd28a-5843-4a74-aa96-ede9032c5f62\",\"total_cost_usd\":0.2000615,\"usage\":{\"input_tokens\":1192,\"cache_creation_input_tokens\":14854,\"cache_read_input_tokens\":50823,\"output_tokens\":806,\"server_tool_use\":{\"web_search_requests\":0,\"web_fetch_requests\":0},\"service_tier\":\"standard\",\"cache_creation\":{\"ephemeral_1h_input_tokens\":14854,\"ephemeral_5m_input_tokens\":0},\"inference_geo\":\"not_available\",\"iterations\":[{\"input_tokens\":2,\"output_tokens\":224,\"cache_read_input_tokens\":14599,\"cache_creation_input_tokens\":255,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":255},\"type\":\"message\"}],\"speed\":\"standard\"},\"modelUsage\":{\"claude-opus-4-8\":{\"inputTokens\":1192,\"outputTokens\":806,\"cacheReadInputTokens\":50823,\"cacheCreationInputTokens\":14854,\"webSearchRequests\":0,\"costUSD\":0.2000615,\"contextWindow\":1000000,\"maxOutputTokens\":64000}},\"permission_denials\":[],\"terminal_reason\":\"completed\",\"fast_mode_state\":\"off\",\"uuid\":\"631ba05e-289d-4471-8909-424461991a14\"}"
+    "{\"type\":\"system\",\"subtype\":\"init\",\"cwd\":\"/sessions/local_2ckhgi8bsn\",\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"tools\":[\"Task\",\"AskUserQuestion\",\"Bash\",\"Edit\",\"Glob\",\"Grep\",\"NotebookEdit\",\"Read\",\"Skill\",\"TaskCreate\",\"TaskGet\",\"TaskList\",\"TaskStop\",\"TaskUpdate\",\"ToolSearch\",\"WebFetch\",\"WebSearch\",\"Write\"],\"mcp_servers\":[],\"model\":\"claude-opus-4-8\",\"permissionMode\":\"default\",\"slash_commands\":[\"deep-research\",\"my-pdf-skill:my-pdf-skill\",\"clear\",\"compact\",\"config\",\"context\",\"heapdump\",\"init\",\"reload-skills\",\"review\",\"security-review\",\"usage\",\"insights\",\"goal\",\"team-onboarding\"],\"apiKeySource\":\"none\",\"claude_code_version\":\"2.1.197\",\"output_style\":\"default\",\"agents\":[\"claude\",\"claude-code-guide\",\"Explore\",\"general-purpose\",\"Plan\",\"statusline-setup\"],\"skills\":[\"deep-research\",\"my-pdf-skill:my-pdf-skill\"],\"plugins\":[{\"name\":\"my-pdf-skill\",\"path\":\"/sessions/local_2ckhgi8bsn/mnt/.local-plugins/marketplaces/local-desktop-app-uploads/my-pdf-skill\",\"source\":\"my-pdf-skill@inline\"}],\"analytics_disabled\":false,\"product_feedback_disabled\":false,\"uuid\":\"97aa4c66-d25d-41e5-9322-3a1e658e5dc7\",\"memory_paths\":{\"auto\":\"/sessions/local_2ckhgi8bsn/mnt/.claude/projects/-sessions-local-2ckhgi8bsn/memory/\"},\"fast_mode_state\":\"off\"}",
+    "{\"type\":\"system\",\"subtype\":\"thinking_tokens\",\"estimated_tokens\":50,\"estimated_tokens_delta\":50,\"uuid\":\"610b4a90-abcc-4c76-a224-c9670e90e94b\",\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\"}",
+    "{\"type\":\"system\",\"subtype\":\"thinking_tokens\",\"estimated_tokens\":138,\"estimated_tokens_delta\":88,\"uuid\":\"2c44e99c-4ae1-413b-8d64-77ecfb8ac350\",\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01ESWwohWenfu2P6f1md7zaG\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"thinking\",\"thinking\":\"\",\"signature\":\"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\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":16350,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":16350},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"e0e6cab3-6ff8-4d31-a7d2-5da6e0569e0a\",\"request_id\":\"req_011CcihszwEGjxG9JfCdHiUY\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01ESWwohWenfu2P6f1md7zaG\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"I'll use the dedicated PDF skill for this since it matches the task exactly. Let me start.\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":16350,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":16350},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"50bb41ff-2098-4694-a06a-d34fa3cb2414\",\"request_id\":\"req_011CcihszwEGjxG9JfCdHiUY\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01ESWwohWenfu2P6f1md7zaG\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01CNTVD5HCViGsHkAgUJ5dpQ\",\"name\":\"Skill\",\"input\":{\"skill\":\"my-pdf-skill:my-pdf-skill\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":16350,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":16350},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"4394aa85-d85f-4d02-b5a4-6bb3b0a7de10\",\"request_id\":\"req_011CcihszwEGjxG9JfCdHiUY\"}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"tool_result\",\"tool_use_id\":\"toolu_01CNTVD5HCViGsHkAgUJ5dpQ\",\"content\":\"Launching skill: my-pdf-skill:my-pdf-skill\"}]},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"0d37a834-50f4-4b2f-b432-2e3a0e4e9f3b\",\"timestamp\":\"2026-07-05T09:52:33.514Z\",\"tool_use_result\":{\"success\":true,\"commandName\":\"my-pdf-skill:my-pdf-skill\"}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"text\",\"text\":\"Base directory for this skill: /sessions/local_2ckhgi8bsn/mnt/.local-plugins/marketplaces/local-desktop-app-uploads/my-pdf-skill/skills/my-pdf-skill\\n\\n# my-pdf-skill\\nPlaceholder skill so the example scenario resolves. Replace with the skill you're testing.\\n\"}]},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"81eebe67-5039-4989-a483-b1da46fecc26\",\"timestamp\":\"2026-07-05T09:52:33.513Z\",\"isSynthetic\":true}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01ESWwohWenfu2P6f1md7zaG\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01AkwRnMPxB68b773Kcmwhfx\",\"name\":\"Bash\",\"input\":{\"command\":\"ls -la /sessions/local_2ckhgi8bsn/mnt/uploads/ 2>/dev/null; echo \\\"---project---\\\"; ls -la /sessions/local_2ckhgi8bsn/mnt/project/ 2>/dev/null\",\"description\":\"List uploads and project folders\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1185,\"cache_creation_input_tokens\":16350,\"cache_read_input_tokens\":0,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":16350},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"cfdcd3af-2f38-4610-be8c-116e25185731\",\"request_id\":\"req_011CcihszwEGjxG9JfCdHiUY\"}",
+    "{\"type\":\"rate_limit_event\",\"rate_limit_info\":{\"status\":\"allowed_warning\",\"resetsAt\":1783317600,\"rateLimitType\":\"seven_day\",\"utilization\":0.85,\"isUsingOverage\":false,\"surpassedThreshold\":0.75},\"uuid\":\"9950e542-4a6b-45db-9a3f-275c9ecd6ed8\",\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\"}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01AkwRnMPxB68b773Kcmwhfx\",\"type\":\"tool_result\",\"content\":\"total 4\\ndrwxr-xr-x 3 ubuntu ubuntu   96 Jul  5 09:52 .\\ndrwxr-xr-x 9 ubuntu ubuntu  288 Jul  5 09:52 ..\\n-rw-r--r-- 1 ubuntu ubuntu 1056 Jul  5 09:52 report.pdf\\n---project---\\ntotal 4\\ndrwxr-xr-x 3 ubuntu ubuntu  96 Jul  5 09:52 .\\ndrwxr-xr-x 9 ubuntu ubuntu 288 Jul  5 09:52 ..\\n-rw-r--r-- 1 ubuntu ubuntu  21 Jul  5 09:52 notes.txt\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"b74c4151-0b03-405f-8d8f-2c3c292248ec\",\"timestamp\":\"2026-07-05T09:52:34.948Z\",\"tool_use_result\":{\"stdout\":\"total 4\\ndrwxr-xr-x 3 ubuntu ubuntu   96 Jul  5 09:52 .\\ndrwxr-xr-x 9 ubuntu ubuntu  288 Jul  5 09:52 ..\\n-rw-r--r-- 1 ubuntu ubuntu 1056 Jul  5 09:52 report.pdf\\n---project---\\ntotal 4\\ndrwxr-xr-x 3 ubuntu ubuntu  96 Jul  5 09:52 .\\ndrwxr-xr-x 9 ubuntu ubuntu 288 Jul  5 09:52 ..\\n-rw-r--r-- 1 ubuntu ubuntu  21 Jul  5 09:52 notes.txt\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01KMY95bx1dhZdBNGk9oQWcm\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"thinking\",\"thinking\":\"\",\"signature\":\"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\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1,\"cache_creation_input_tokens\":1993,\"cache_read_input_tokens\":16350,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1993},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"293e7b8e-f1eb-4ebc-8f29-78fb68874737\",\"request_id\":\"req_011CcihtUSCZCXUdn4Y8dGQ9\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01KMY95bx1dhZdBNGk9oQWcm\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"The skill is a placeholder, so I'll proceed with the task directly. Let me read the PDF.\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1,\"cache_creation_input_tokens\":1993,\"cache_read_input_tokens\":16350,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1993},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"c7187d57-5920-4c2f-b9f5-a3454133d0b2\",\"request_id\":\"req_011CcihtUSCZCXUdn4Y8dGQ9\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01KMY95bx1dhZdBNGk9oQWcm\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01QDXZpwCiBoBkRRwWiSYv3Z\",\"name\":\"Read\",\"input\":{\"file_path\":\"/sessions/local_2ckhgi8bsn/mnt/uploads/report.pdf\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":1,\"cache_creation_input_tokens\":1993,\"cache_read_input_tokens\":16350,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1993},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"7fd5d38c-723f-4e38-b83a-b150791dda2f\",\"request_id\":\"req_011CcihtUSCZCXUdn4Y8dGQ9\"}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01QDXZpwCiBoBkRRwWiSYv3Z\",\"type\":\"tool_result\",\"content\":\"PDF file read: /sessions/local_2ckhgi8bsn/mnt/uploads/report.pdf (1KB)\"}]},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"bcf5c078-7637-4ee9-939d-25be541d5e0b\",\"timestamp\":\"2026-07-05T09:52:38.805Z\",\"tool_use_result\":{\"type\":\"pdf\",\"file\":{\"filePath\":\"/sessions/local_2ckhgi8bsn/mnt/uploads/report.pdf\",\"base64\":\"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\",\"originalSize\":1056}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"document\",\"source\":{\"type\":\"base64\",\"media_type\":\"application/pdf\",\"data\":\"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\"}}]},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"aa33c287-e709-41ad-bef6-4110843f6633\",\"timestamp\":\"2026-07-05T09:52:38.800Z\",\"isSynthetic\":true}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01WFLkoXaKXUhuiLwn34ozK3\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"thinking\",\"thinking\":\"\",\"signature\":\"EuECCmMIDxgCKkBgZbRkVqyblTGmoFqhQq9lyfVfCvHpIrPctPEsDO+c4M3PE8nQ22F71xQxgLSrAfrv0V7w+7PVhKSupnjT3qHeMg9jbGF1ZGUtb3B1cy00LTg4AEIIdGhpbmtpbmcSDLBpIbibWwGKFj3rZBoMtRG5pGWgEA9EeOr1IjB1g9fuh/Hdah6bFdTGtslbzm12kCBE2VKOzsnnfE0cHtaGyqZO6otY565r012YWGwqqwGrk8UfN0XeOtRAxVcVr04FnmnWRfC6sp7aGVenkJTF69IV+IhdD6oKRJGvpEcOtJ1REDMkiX1F66b3fuHJcb/wDGsHJ0IsXrd3xx93KeOvuxmD3iyunTXd8dtDWpnaYDQdmSjKyk/kobpabRBNCe381eemZBxq8uqXKmHyW0szaNH/DKIoizbY5WI6AtVR0nzfCZQBTdwB/5ziJRcxNGqInF0eY9um0pDAr88YAQ==\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":1955,\"cache_read_input_tokens\":18343,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1955},\"output_tokens\":5,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"1d8c5d19-1730-4476-b071-1704157e204b\",\"request_id\":\"req_011Ccihtmz8pyMu9AJamFFyb\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01WFLkoXaKXUhuiLwn34ozK3\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"I have the report. Let me write the action items checklist to the outputs folder.\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":1955,\"cache_read_input_tokens\":18343,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1955},\"output_tokens\":5,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"874fe646-32f1-4df6-8e09-9720d17303fa\",\"request_id\":\"req_011Ccihtmz8pyMu9AJamFFyb\"}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_01WFLkoXaKXUhuiLwn34ozK3\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01JALx8ysLPTcwFDqpnZMweX\",\"name\":\"Write\",\"input\":{\"file_path\":\"/sessions/local_2ckhgi8bsn/mnt/project/outputs/actions.md\",\"content\":\"# Q2 Project Status Report — Action Items\\n\\n- [ ] **Alice:** Finalize the production rollout plan by June 20.\\n- [ ] **Bob:** Add monitoring dashboards for the new pipeline.\\n- [ ] **Carol:** Update the on-call runbook and notify support.\\n\"},\"caller\":{\"type\":\"direct\"}}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":1955,\"cache_read_input_tokens\":18343,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":1955},\"output_tokens\":5,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"7193d35b-ab7b-4a2a-9bff-6d3ed0dc2cf5\",\"request_id\":\"req_011Ccihtmz8pyMu9AJamFFyb\"}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01JALx8ysLPTcwFDqpnZMweX\",\"type\":\"tool_result\",\"content\":\"File created successfully at: /sessions/local_2ckhgi8bsn/mnt/project/outputs/actions.md (file state is current in your context — no need to Read it back)\"}]},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"996d3b1e-f5c1-4a4a-86b8-6b4ccb61ad55\",\"timestamp\":\"2026-07-05T09:52:44.239Z\",\"tool_use_result\":{\"type\":\"create\",\"filePath\":\"/sessions/local_2ckhgi8bsn/mnt/project/outputs/actions.md\",\"content\":\"# Q2 Project Status Report — Action Items\\n\\n- [ ] **Alice:** Finalize the production rollout plan by June 20.\\n- [ ] **Bob:** Add monitoring dashboards for the new pipeline.\\n- [ ] **Carol:** Update the on-call runbook and notify support.\\n\",\"structuredPatch\":[],\"originalFile\":null,\"userModified\":false}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-4-8\",\"id\":\"msg_017XuKKAtw1jNYVyu1b7RTtn\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"The report is a Q2 Project Status Report on the ingest pipeline migration. In short: the migration is on track, end-to-end latency dropped 30% after the caching layer landed, and error rates are within target. The one remaining risk called out is the data-backfill window. It closes with three owner-assigned action items.\\n\\nI saved those as a checklist here: [View actions.md](computer:///sessions/local_2ckhgi8bsn/mnt/project/outputs/actions.md)\\n\\n- Alice — finalize the production rollout plan by June 20\\n- Bob — add monitoring dashboards for the new pipeline\\n- Carol — update the on-call runbook and notify support\"}],\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":334,\"cache_read_input_tokens\":20298,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":334},\"output_tokens\":1,\"service_tier\":\"standard\",\"inference_geo\":\"not_available\"},\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"uuid\":\"2d8a88b5-edf6-453a-91ce-df31db767377\",\"request_id\":\"req_011CcihuA4SCpG4JvaqvRd3J\"}",
+    "{\"type\":\"result\",\"subtype\":\"success\",\"is_error\":false,\"api_error_status\":null,\"duration_ms\":19763,\"duration_api_ms\":19025,\"ttft_ms\":3621,\"ttft_stream_ms\":2071,\"time_to_request_ms\":32,\"num_turns\":7,\"result\":\"The report is a Q2 Project Status Report on the ingest pipeline migration. In short: the migration is on track, end-to-end latency dropped 30% after the caching layer landed, and error rates are within target. The one remaining risk called out is the data-backfill window. It closes with three owner-assigned action items.\\n\\nI saved those as a checklist here: [View actions.md](computer:///sessions/local_2ckhgi8bsn/mnt/project/outputs/actions.md)\\n\\n- Alice — finalize the production rollout plan by June 20\\n- Bob — add monitoring dashboards for the new pipeline\\n- Carol — update the on-call runbook and notify support\",\"stop_reason\":\"end_turn\",\"session_id\":\"9fefa516-b19c-48f5-b636-97381a0e56d9\",\"total_cost_usd\":0.2657655,\"usage\":{\"input_tokens\":1190,\"cache_creation_input_tokens\":20632,\"cache_read_input_tokens\":54991,\"output_tokens\":1040,\"server_tool_use\":{\"web_search_requests\":0,\"web_fetch_requests\":0},\"service_tier\":\"standard\",\"cache_creation\":{\"ephemeral_1h_input_tokens\":20632,\"ephemeral_5m_input_tokens\":0},\"inference_geo\":\"not_available\",\"iterations\":[{\"input_tokens\":2,\"output_tokens\":215,\"cache_read_input_tokens\":20298,\"cache_creation_input_tokens\":334,\"cache_creation\":{\"ephemeral_5m_input_tokens\":0,\"ephemeral_1h_input_tokens\":334},\"type\":\"message\"}],\"speed\":\"standard\"},\"modelUsage\":{\"claude-opus-4-8\":{\"inputTokens\":1190,\"outputTokens\":1040,\"cacheReadInputTokens\":54991,\"cacheCreationInputTokens\":20632,\"webSearchRequests\":0,\"costUSD\":0.2657655,\"contextWindow\":1000000,\"maxOutputTokens\":64000}},\"permission_denials\":[],\"terminal_reason\":\"completed\",\"fast_mode_state\":\"off\",\"uuid\":\"ea4d1b1b-4c21-47d8-b5ff-c0a4593758f8\"}"
   ],
   "controlOut": [
-    "{\"type\":\"control_request\",\"request_id\":\"init-1\",\"request\":{\"subtype\":\"initialize\",\"appendSubagentSystemPrompt\":\"## Cowork environment\\n\\nYou are running as a subagent inside a Cowork session. Shell commands execute in an isolated Linux sandbox rooted at /sessions/local_9hldogcxp — files created there (or under /tmp) exist only in the sandbox, not on the user's real computer. User-attached folders are mounted under /sessions/local_9hldogcxp/mnt/.\",\"hooks\":{\"PreToolUse\":[{\"matcher\":\"Task\",\"hookCallbackIds\":[\"cowork-task-bg-block\"]}]}}}",
+    "{\"type\":\"control_request\",\"request_id\":\"init-1\",\"request\":{\"subtype\":\"initialize\",\"appendSubagentSystemPrompt\":\"## Cowork environment\\n\\nYou are running as a subagent inside a Cowork session. Shell commands execute in an isolated Linux sandbox rooted at /sessions/local_2ckhgi8bsn — files created there (or under /tmp) exist only in the sandbox, not on the user's real computer. User-attached folders are mounted under /sessions/local_2ckhgi8bsn/mnt/.\",\"hooks\":{\"PreToolUse\":[{\"matcher\":\"Task\",\"hookCallbackIds\":[\"cowork-task-bg-block\"]}]}}}",
     "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"text\",\"text\":\"Read report.pdf from your uploads, summarize it, and write the action items\\nto outputs/actions.md as a markdown checklist.\\n\"}]}}"
   ],
   "effectiveFidelity": "container",
@@ -89,11 +92,14 @@
     "outputs",
     "project"
   ],
+  "preRunPaths": [
+    "project/notes.txt"
+  ],
   "scenarioSource": "../scenarios/example-pdf-skill.yaml",
   "fingerprint": {
     "baseline": "1.18286.0",
-    "skillHash": "4db1ddd5d109abc71f9b28807d3e31729f9882d9b85d6d0a9dd7a1980bed3110",
-    "contentSig": "e8022822156176fa10e04f0e31d1269719fd09890488db3dfd6351ea8c34431c",
+    "skillHash": "b760ea90682777367fbd44d866ea66d316452da1abb18bf8cb61f1cec8e67806",
+    "contentSig": "ddf68cf6d74e0599729589f8be4f0a9eec440093c99c1060cd6f47dd671a0dfd",
     "skillSources": [
       "../skills/my-pdf-skill"
     ],

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "cowork-harness",
-  "version": "0.25.0",
+  "version": "0.26.0",
   "description": "Scriptable, CI-friendly harness for Claude Cowork's runtime contract for testing skills across scenarios — same agent, mounts, egress allowlist, permission protocol, and sandbox limitations.",
   "license": "MIT",
   "type": "module",

package/schema/cassette.v8.json

@@ -0,0 +1,166 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "$id": "https://raw.githubusercontent.com/yaniv-golan/cowork-harness/main/schema/cassette.v8.json",
+  "title": "cowork-harness cassette v8",
+  "description": "Recorded control-protocol cassette produced by cowork-harness (https://github.com/yaniv-golan/cowork-harness). Contains a frozen scenario, recorded agent event stream, control-protocol responses, artifact manifest, and a staleness fingerprint. v3 adds `contentSig`. v4 persists `userVisibleRoots`. v6 (staleness redesign, breaking): contentSig unified onto the skillHash walk (same file set incl. OS-junk exclusion, in-tree-symlink-by-target); git-tracked file set is the DEFAULT boundary (opt out with COWORK_HARNESS_GITSET=0). v7: NUL-byte entry separator in hashDir and computeContentSig to prevent hash collisions for file paths containing newline characters. v8: skillHash folds fixed-length content shas (not raw bytes) and contentSig/link entries are type-prefixed & NUL-framed (closes unframed-concatenation collisions → a v8 staleness hash is non-comparable with pre-v8, so `rehash` routes pre-v8 cassettes to re-record); `ManifestEntry.truncationReason` names why a body is absent (the cassette-level `readonlyFolderRoots` list is removed).",
+  "type": "object",
+  "required": [
+    "scenario",
+    "events"
+  ],
+  "properties": {
+    "$schema": {
+      "type": "string",
+      "format": "uri"
+    },
+    "generator": {
+      "type": "string",
+      "const": "cowork-harness"
+    },
+    "cassetteVersion": {
+      "type": "integer",
+      "minimum": 0
+    },
+    "scenario": {
+      "type": "object",
+      "description": "Embedded scenario snapshot at record time."
+    },
+    "events": {
+      "type": "array",
+      "items": {
+        "type": "string"
+      },
+      "description": "Recorded agent event stream (newline-delimited JSON lines)."
+    },
+    "controlOut": {
+      "type": "array",
+      "items": {
+        "type": "string"
+      },
+      "description": "Driver→agent control_response stream (newline-delimited JSON lines)."
+    },
+    "effectiveFidelity": {
+      "type": "string",
+      "description": "The fidelity tier the live record resolved to."
+    },
+    "artifacts": {
+      "type": "array",
+      "description": "Snapshot of user-visible output files produced by the run (the manifest). Each entry carries path + bytes + sha256; small text/JSON bodies are inlined, larger ones are hash-only.",
+      "items": {
+        "type": "object",
+        "required": ["path", "bytes", "sha256"],
+        "properties": {
+          "path": { "type": "string", "description": "workRoot-relative path, e.g. outputs/report.json" },
+          "bytes": { "type": "number", "description": "file size in bytes at record time" },
+          "sha256": { "type": "string", "description": "sha256 of the raw bytes (empty string only on an unreadable/containment-failed entry)" },
+          "body": { "type": "string", "description": "inlined body (≤ the artifact-body cap); materialized on replay so artifact_json can parse it. Absent on a truncated entry." },
+          "encoding": { "type": "string", "enum": ["utf8", "base64"], "description": "how `body` is encoded (base64 for non-UTF-8/binary)." },
+          "truncated": { "type": "boolean", "description": "body omitted (hash-only). file_exists/user_visible_artifact still PASS (existence proven by path+sha); artifact_json is evidence-unavailable." },
+          "truncationReason": {
+            "type": "string",
+            "enum": ["size", "readonly", "unreadable"],
+            "description": "v8+: WHY the body is absent, so replay gives the precise artifact_json remedy. `size` = over the body cap (raise --max-artifact-bytes); `readonly` = a mode:r connected-folder input (assert on a deliverable); `unreadable` = a record-time read/containment failure (sha256 is \"\"). ABSENT on pre-v8 entries."
+          }
+        }
+      }
+    },
+    "userVisibleRoots": {
+      "type": "array",
+      "items": {
+        "type": "string"
+      },
+      "description": "v4+: the user-visible mount roots captured at record time (`outputs` + each connected folder's resolved mount name). Replay reads THIS instead of a hardcoded prefix. ABSENT on pre-v4 cassettes → replay falls back to the legacy prefix."
+    },
+    "preRunPaths": {
+      "type": "array",
+      "items": {
+        "type": "string"
+      },
+      "description": "workRoot-relative paths under userVisibleRoots that existed BEFORE the agent ran — the baseline `no_unexpected_files` diffs against on replay. Optional metadata (no cassetteVersion bump); ABSENT on pre-0.24 and microvm recordings → replay EXCLUDES `no_unexpected_files` with a loud warning. Always co-present with `userVisibleRoots` when present."
+    },
+    "scenarioSource": {
+      "type": "string",
+      "description": "The authored scenario SOURCE file this cassette was recorded from, RELATIVE to the cassette dir (relocatable, no absolute host path). `record --rerecord-stale` prefers this over a slug guess. ABSENT when recorded from an in-memory/inline scenario."
+    },
+    "fingerprint": {
+      "type": "object",
+      "description": "Staleness tripwire over the skill/baseline inputs that determined the recording.",
+      "properties": {
+        "baseline": {
+          "type": "string"
+        },
+        "skillHash": {
+          "type": "string",
+          "description": "Algorithm-specific hash of skill dir contents (staleness detection)."
+        },
+        "contentSig": {
+          "type": "string",
+          "description": "v3+: Algorithm-independent SHA-256 over sorted relpath:content-sha256 pairs. Used by `rehash` to verify content is unchanged when the staleness algorithm changes."
+        },
+        "skillSources": {
+          "type": "array",
+          "items": {
+            "type": "string"
+          }
+        },
+        "skillScope": {
+          "type": "array",
+          "items": {
+            "type": "string"
+          }
+        },
+        "sharedHash": {
+          "type": "string"
+        },
+        "fileSigs": {
+          "type": "array",
+          "description": "v5+: per-file manifest [relpath, contentSha] of files feeding skillHash, for exact-diff staleness reporting. Root-relative paths; scanned+redacted like skillSources.",
+          "items": {
+            "type": "array",
+            "items": {
+              "type": "string"
+            },
+            "minItems": 2,
+            "maxItems": 2
+          }
+        },
+        "fileSigsOmitted": {
+          "type": "boolean",
+          "description": "v5+: true when the per-file manifest was omitted because the tree exceeded the size cap (degrades to bucket-level staleness messages)."
+        },
+        "mode": {
+          "type": "string",
+          "enum": [
+            "git",
+            "raw"
+          ],
+          "description": "Phase C: the file-set boundary used for skillHash — 'git' (git-tracked set — the DEFAULT unless COWORK_HARNESS_GITSET=0, and every dir is a git work tree) or 'raw' (filesystem walk; used when GITSET=0 OR any dir is not a repo). A record-vs-verify mode flip means hashes aren't comparable → re-record."
+        },
+        "agentScope": {
+          "type": "string",
+          "enum": ["skill"],
+          "description": "Opt-in per-skill agent scoping (COWORK_HARNESS_AGENT_SCOPE=skill) was active for this scoped hash: a skill-named agents/<n> was treated as skill <n>'s private input, not a fleet-wide shared root. ABSENT = the default (agents/ is shared). A record-vs-verify flip → re-record under the same setting."
+        }
+      }
+    },
+    "authoring": {
+      "type": "object",
+      "description": "Provenance of how the gate answers were authored. PRESENT with nonDeterministic:true only when a live decider answered at least one gate during recording (a driving agent via --decider-dir, a model via --decider-llm, or an --on-unanswered first auto-pick), so RE-recording may drift. The cassette itself still replays deterministically (the answers are frozen). ABSENT means fully scripted/deterministic authoring.",
+      "properties": {
+        "nonDeterministic": {
+          "type": "boolean"
+        },
+        "channel": {
+          "type": "string",
+          "enum": [
+            "decider-dir",
+            "decider-llm"
+          ]
+        }
+      },
+      "required": [
+        "nonDeterministic"
+      ]
+    }
+  }
+}

package/schema/run-result.json

@@ -41,7 +41,7 @@
     "egress": { "type": "array" },
     "assertions": { "type": "array" },
     "subagents": { "type": "array" },
-    "unanswered": { "type": "array" },
+    "nonReproducibleAnswers": { "type": "array" },
     "usage": {
       "type": "object",
       "description": "SDK usage payload (input_tokens, output_tokens, etc.), pass-through, plus a harness-computed 'turns' field (from the SDK result message's num_turns) when available. Wave 0 seam.",
@@ -58,10 +58,50 @@
       }
     },
     "durationMs": { "type": "number" },
+    "fingerprint": {
+      "type": "object",
+      "description": "Skill/plugin staleness fingerprint recorded at run time; persisted so `verify-run` can detect a kept run that predates a skill change and refuse to vouch for answer-coverage against stale gate labels.",
+      "required": ["baseline"],
+      "properties": {
+        "baseline": { "type": "string", "description": "appVersion at record time" },
+        "skillHash": { "type": "string", "description": "hash of the session's local skill/plugin/marketplace dir contents (if any)" },
+        "skillSources": { "type": "array", "items": { "type": "string" } },
+        "skillScope": { "type": "array", "items": { "type": "string" } },
+        "sharedHash": { "type": "string", "description": "shared-root hash for scoped cassettes; absent on whole-tree or non-plugin-root mounts" },
+        "contentSig": { "type": "string", "description": "algorithm-independent content fingerprint; used by `rehash` to verify content is unchanged across format bumps" },
+        "fileSigs": {
+          "type": "array",
+          "description": "per-file manifest [relpath, contentSha] of the exact files feeding skillHash",
+          "items": { "type": "array", "items": { "type": "string" }, "minItems": 2, "maxItems": 2 }
+        },
+        "fileSigsOmitted": { "type": "boolean" },
+        "mode": {
+          "type": "string",
+          "enum": ["git", "raw"],
+          "description": "the file-set boundary used for skillHash — \"git\" (git-tracked set — the DEFAULT for a git work tree unless COWORK_HARNESS_GITSET=0) or \"raw\" (filesystem walk; used when GITSET=0 or the dir is not a git work tree)"
+        },
+        "agentScope": { "type": "string", "enum": ["skill"] }
+      }
+    },
     "outDir": { "type": "string" },
     "workDir": { "type": "string" },
     "outputsDir": { "type": "string" },
+    "userVisibleRoots": {
+      "type": "array",
+      "items": { "type": "string" },
+      "description": "The user-visible mount roots (relative to mnt/) for this run — `outputs` plus each connected work folder's resolved mount name. Plugins are not included (read-only inputs, never artifact roots)."
+    },
+    "readonlyFolderRoots": {
+      "type": "array",
+      "items": { "type": "string" },
+      "description": "Subset of userVisibleRoots that are read-only (mode:\"r\") connected-folder mounts — inputs, not deliverables. `artifacts` excludes them so `scaffold` doesn't emit file_exists for an input."
+    },
     "artifacts": { "type": "array" },
+    "preRunPaths": {
+      "type": "array",
+      "items": { "type": "string" },
+      "description": "workRoot-relative paths that existed under the user-visible roots BEFORE the agent ran (captured post-staging, pre-spawn) — the baseline `no_unexpected_files` diffs against. Absent when the tier didn't capture it (microvm) or the run predates the seam."
+    },
     "partial": { "type": "boolean" },
     "unansweredGate": { "type": "object" },
     "nonDeterministic": { "type": "boolean" },
@@ -144,6 +184,20 @@
         "full": { "type": "number" },
         "partial": { "type": "number" }
       }
+    },
+    "toolResults": {
+      "type": "array",
+      "description": "Tool-result text at assertion-fidelity cap (10 KB per result). Used by tool_result_contains / tool_result_not_contains. `assertText` is preferred when present; falls back to `text` (500-char display cap) for cassettes recorded before this field was added.",
+      "items": {
+        "type": "object",
+        "required": ["isError", "text"],
+        "properties": {
+          "toolUseId": { "type": "string" },
+          "isError": { "type": "boolean" },
+          "text": { "type": "string" },
+          "assertText": { "type": "string" }
+        }
+      }
     }
   }
 }