@sleepy_panda_srl/portulan 0.3.0 → 0.4.0

package/core/personas/README.md

@@ -29,10 +29,15 @@ token; `effort`, one of `low`, `medium`, `high`, `xhigh` or `max`; `omitClaudeMd
 unquoted, whether the role runs without the project's instruction files; and `cacheLifetime`, `5m` or
 `1h`. The effort levels and the name `omitClaudeMd` are adopted from Claude Code's vocabulary, the first
 host to express them. `effort` sets the host subagent's reasoning effort, and is not the thinking cap the
-`tools/local-llm` runner's `--effort` sends with each request it makes. `doctor` fails a
-malformed value and `compile` refuses one, and neither supplies a value the persona did not write. For a
-persona a composed pack ships, `compile` writes the declared fields into the frontmatter of the role's
-hand-written host binding, `agents/<name>.md` in the workspace's tree, and leaves every other line of
+`tools/local-llm` runner's `--effort` sends with each request it makes. `doctor` fails malformed values
+in composed pack personas; `compile` validates every persona source it compiles. Neither supplies a value
+the persona did not write. `compile` writes declared fields from composed pack personas into the
+frontmatter of the role's existing host binding: `agents/<name>.md` in a plugin tree and
+`.claude/agents/<name>.md` in an ordinary project. Compiler and doctor share that selection; a malformed,
+unreadable or escaping plugin marker is not treated as an ordinary project. When
+the declared tree resolves to this compiler's own engine root, including a canonical alias, the same
+pipeline also reads `core/personas/*.md` except this README. A foreign tree's matching files do not
+participate. A core source's declared name must match its filename. The compiler leaves every other line of
 that file as its author wrote it; `--check` byte-compares the result, and a line inside the compiled
 block that `compile` did not write is refused rather than removed. One it cannot catch: a line added inside
 the block in exactly the spelling and order `compile` writes reads as compiled, and is dropped once the
@@ -41,19 +46,56 @@ field, `compile` owns every top-level key it could write, so a hand-written `exp
 even for a key other than `cacheTtl`. A field a host cannot express is
 named, with the persona and the host, in the degradation report `compile` prints. Claude Code takes all
 four on an agent (`cacheLifetime` as `experimental.cacheTtl`), and a vendored `AGENTS.md` takes none.
-**Claude Code loads `agents/` only from a plugin root**, and a repository that is not a plugin loads its
-agents from `.claude/agents/`, which `compile` does not write. So outside a plugin the compiled block is
-loaded by nothing, and the degradation report says so for each persona it applies to. Where the tree is
-a plugin root the report is silent, which is not proof of loading: a plugin's agents load only while it
-is installed and enabled, and `compile` cannot see that. _(Provenance:
+`compile` writes only existing bindings; an absent adopter binding remains unbound and is never
+created by compilation. A persona-only workspace can compile its declared fields without a gate
+policy or guidance. Nonregular, escaping and invalid UTF-8 inputs are refused before writes.
+Doctor checks the selected binding's name and declared tool grant, keeping an unread file distinct
+from an absent one. Plugin packaging checks core and actually composed charter owners in both
+directions, including personas declaring no context fields. File equality proves neither plugin
+enablement nor that a running host honored a field. _(Provenance:
 proposal `0036`'s order of work, item 7, and proposal `0039`'s first two rules. The host's keys, values
 and agent locations were read in Claude Code 2.1.317's program text on 2026-11-03.)_
 
-A binding has one persona owner across composed packs, including an unbound persona or one declaring
+This product binds both composed roles. Supervisor still requires a fresh invocation for a checkpoint.
+Local-worker's host wrapper points to its existing loopback runner; loading that role does not itself
+invoke a local model or prove local execution.
+
+The three core sources explicitly declare `model: inherit`, `omitClaudeMd: false` and
+`cacheLifetime: 5m`, leaving effort undeclared. These are product-role policy choices: inherit the
+parent model, retain the instruction floor and request a five-minute cache lifetime. In Claude Code,
+explicit `inherit` outranks the ordinary `CLAUDE_CODE_SUBAGENT_MODEL` default; an invocation's model
+selection and `CLAUDE_CODE_SUBAGENT_MODEL_FORCE=1` can override it. Cache lifetime remains subject to
+the host's precedence rules. Compilation proves the committed bytes, with no claim about quality,
+savings or live loading. _(Host precedence checked against the
+[official subagent documentation](https://code.claude.com/docs/en/sub-agents) on 2026-11-09.)_
+
+The acquired [implementer](../../evals/ab/personas/implementer.md),
+[reviewer](../../evals/ab/personas/reviewer.md) and
+[librarian](../../evals/ab/personas/librarian.md) comparisons each have three
+frozen artifact cases, one attempt per arm, through native children on Sonnet 5.5
+in Claude Code 2.1.321. Both arms pass all nine cases. Medium has a lower
+cold-normalized cost index in this observation; observed cache costs differ,
+and the reviewer arm's observed cost rises. These results support the retained
+policy on that corpus and model, with no global medium default earned for other
+inherited models. The [protocol](../../evals/ab/personas/README.md) discloses the
+small sample, both probes, full parent/child usage and one retained-answer
+decoder repair after the first outcome. The shipped offline checker regrades
+the records and checks routing and source bytes.
+
+A binding has one persona owner across core and composed pack sources, including an unbound persona or one declaring
 no fields; duplicate destinations are refused before any artifact is written. The binding must be a
 regular file whose real path stays in the real declared tree, including its parent directories.
 `omitClaudeMd` accepts only unquoted `true` or `false`, in both declarations and compiled blocks.
 
+`upgrade` refreshes existing eligible bindings through the same compiler planner,
+including normal host-cache pack discovery and its shadow refusals. It preserves
+unrelated binding text and modes, creates no absent binding, and writes no gate
+settings. Core sources still participate only in the installed engine's own
+declared tree; an ordinary adopter receives the baked core agents by updating the
+plugin, without installing a foreign core adapter. The command-level historical
+v0.2.0 demonstration is in
+[`../../cli/persona-upgrade.test.mjs`](../../cli/persona-upgrade.test.mjs).
+
 ## Status
 
 **Milestone 3 — contract, exemplars, and one host binding.** This README fixes the persona contract; the

package/core/personas/implementer.md

@@ -2,6 +2,9 @@
 name: implementer
 description: The default working persona. Takes one task through the full loop — research → plan → implement → verify → learn — and hands back a verified change plus what it learned. Delegate a unit of implementation work to it.
 tools: [read, search, edit-code, run-verify, draft-memory, open-pr]
+model: inherit
+omitClaudeMd: false
+cacheLifetime: 5m
 ---
 
 # Persona — Implementer

package/core/personas/librarian.md

@@ -2,6 +2,9 @@
 name: librarian
 description: The persona that tends the curated layer — consolidating memory, regenerating the size-budgeted index, drafting rule-change proposals from incidents and PR reviews, and chasing stale rules toward retirement. It drafts and files; it never accepts. Delegate a memory/evolution housekeeping pass to it.
 tools: [read, search, read-history, draft-memory, draft-proposal, write-index, open-pr]
+model: inherit
+omitClaudeMd: false
+cacheLifetime: 5m
 ---
 
 # Persona — Librarian

package/core/personas/reviewer.md

@@ -2,6 +2,9 @@
 name: reviewer
 description: The last-mile check on a change before it merges. Reads a proposed diff against the task's acceptance criteria and the definition of done, re-runs the verify recipe, and returns a verdict — it does not edit the code or merge. Delegate review of a completed change to it.
 tools: [read, search, run-verify, post-review]
+model: inherit
+omitClaudeMd: false
+cacheLifetime: 5m
 ---
 
 # Persona — Reviewer

package/core/templates/handoff.md

@@ -10,12 +10,16 @@
 > decisions and their rationale. When one is owed, and why the commit carries the rest:
 > `../operating/loop.md`.)_
 >
+> Link the records that own measured figures and durable rules. Keep each open decision and its why.
+> A declared token cap binds new handoffs; compress or move and link facts, never split or drop a why.
+> The heading names the outcome; the filename already carries the date. See `../operating/context.md`.
+>
 > **The commit a change lands in:** a subject that says what changed, in the present tense and under 72
 > characters; a blank line; a body of a few lines that says why, and what was rejected when that matters.
 
 ---
 
-# Handoff — {task / thread}
+# {outcome / open task}
 
 **State.** {where things stand: what is done and committed, what is in progress and uncommitted.}
 
@@ -30,7 +34,7 @@
 
 _A filled one, for a session that stopped mid-change:_
 
-# Handoff — retry the flaky upload test
+# Retry the flaky upload test
 
 **State.** The retry wrapper is written and its unit test passes; it is stashed, not committed, because
 the integration suite still fails once in about ten runs.

package/evals/ab/personas/README.md

@@ -0,0 +1,111 @@
+# Core persona routing comparisons
+
+This protocol addresses [#520](https://github.com/sleepy-panda-srl/portulan/issues/520).
+The frozen [corpus](corpus.json) has three artifact tasks per core role and one
+attempt per arm. The [grader](grader.mjs) checks semantic behavior: bounded
+execution and nonmutation for implementer code, executable counterexamples and
+a clean control for reviewer findings, and preservation, escalation and human
+authority for librarian drafts. These observations do not establish general
+quality or a reliability rate.
+
+A is the shipped role: inherited model, undeclared effort, instruction floor
+retained and five-minute cache lifetime. B requests medium effort on the same
+concrete inherited model through one native foreground persona child. The parent
+runs at high effort; host request metadata must establish effective child effort.
+The `--agent` main-session override is not this mechanism. Each parent and child
+has a fresh context. Case order alternates A/B then B/A; an invalid invocation or
+failed semantic control stops acquisition without resampling.
+
+The observed concrete model is Sonnet 5.5 on Claude Code 2.1.321. A medium win on
+this model and corpus does not establish medium effort for other inherited
+models. The product retains `model: inherit` and undeclared effort; no global
+lower-effort default is earned here.
+
+## Acquisition and accounting
+
+Private originals retain the journal, CLI output, parent/child transcripts and
+frozen source snapshots. Publication whitelists generic artifact answers,
+numeric request metadata, normalized command arguments, timestamps and digests.
+Answer presentation replaces only known acquisition and plugin-root prefixes
+with `<acquisition>` and `<plugin>`; the record lists the affected answer and
+retains both original and presented answer digests. Artifact contents and scores
+are unchanged. Original answer wording otherwise remains.
+Request identities and context identities are deterministically hashed; original
+local paths, session identities and unrelated conversation content do not ship.
+Public digests permit integrity checks; fidelity to private originals additionally
+requires the independent acquisition review.
+
+Acquisition fidelity **PASS**, signed `/root/checkpoint_520_open`, independent
+read-only reviewer, 2026-11-10, for capture
+`ff3688658bc20af6b78e834a5149ea6bd52bc88027b0340ec98cb52430c6d641`.
+The reviewer independently checked all 80 requests (39 parent, 41 child), native
+invocations, frozen sources, original answers, repair history and cost arithmetic.
+The one disclosed citation normalization matches the deterministic transformation
+and changes no JSON artifact or score. The checker and four capture/package tests
+passed. This establishes retained-event fidelity; the sample, decoder repair,
+reconstructed first-probe command and absence of billing evidence remain limits.
+
+Two mechanism probes precede the eighteen quality children. The first used the
+host's then-default Fable 5.2 and exceeded the configured one-dollar host-estimate
+ceiling after a child ran. Its failed parent and child requests remain overhead;
+it earns no quality result. Its command was reconstructed after execution from
+the retained source, which was also snapshotted after execution. It is not a
+contemporaneously recorded command. The second probe selected Sonnet 5.5 before
+any quality outcome and demonstrated the native medium-effort mechanism. Quality
+inference is limited to that model.
+
+Every observed parent, child and probe request is counted once. The frozen price
+vector is an index in uncached-input-token equivalents, not a price quotation or
+subscription bill: uncached input 1, five-minute writes 1.25, one-hour writes 2,
+unclassified writes 1.25, reads 0.1 and output 5. Registers separate observed
+cost from a cold-normalized index. Normalization reprices all input in the first
+request of each fresh parent and child context at 1.25, retaining subsequent
+requests' observed classes and all output. Probe overhead is reported separately
+from role comparisons. CLI USD values are host lifecycle estimates; they are not
+added to these indexes.
+
+## Disclosed decoder repair
+
+After the first quality outcome, acquisition stopped because the original
+decoder rejected a correct function inside one JSON fence followed by prose.
+The task requested a JSON artifact without forbidding surrounding prose; the
+original decoder already allowed a whole-answer fence. The retained answer
+passed the unchanged semantic execution checks.
+
+The [original grader](repairs/grader-before-decode-repair.mjs.txt) and
+[first answer](repairs/first-control-answer.txt) remain. A bounded decoder repair
+accepts exactly one JSON fence with prose and rejects competing artifacts or
+ambiguous fences. Semantic predicates, cases and pricing code stayed byte for
+byte identical. The original false score is preserved alongside the repaired
+true score. An independent checkpoint approved the exact decoder and continuation
+instrument hashes before the remaining seventeen untouched calls. The first
+child is never rerun. This is a post-outcome instrument repair, not an unchanged
+preregistration.
+
+## Offline verification and delivery
+
+The offline checker is `node cli/persona-routing.mjs --repo-root .`; it never
+starts a host. It checks the immutable acquisition digest, all three role records,
+semantic scores, generated registers, complete case/arm coverage, source and
+binding bytes, effective effort, unique contexts/requests, frozen task bytes,
+protocol snapshots, repair history and usage accounting. Exit 1 is a finding;
+exit 2 is unavailable data or a failed read. The existing A/B recipe runs it.
+
+The records and costs remain observations of source `01ca4887`; later bundle
+changes are not new measurements. Current core charters must remain byte-identical
+to that capture. A current plugin binding may change only its one nonempty,
+single-line description scalar: every other byte, including tools, body and
+context declarations, must match the historical binding. Malformed, duplicate,
+missing or multiline descriptions refuse comparison. This permits the later
+librarian description compression while preserving the acquired routing contract;
+it establishes neither current whole-bundle costs nor equivalent automatic
+delegation after description changes.
+
+The npm payload carries the records, frozen protocol evidence, grader and offline
+checker. Live acquisition and private normalization commands remain workshop
+tools. A plugin update delivers the baked `agents/` files. The CLI's persona form
+migration refreshes existing eligible bindings through the compiler planner:
+core sources only in the installed engine's own tree, composed pack sources
+through normal pack resolution. It creates no absent binding, installs no
+foreign core adapter and changes no gate settings. The historical v0.2.0 suite
+demonstrates actual check/write/check, idempotence, refusal and rollback.

package/evals/ab/personas/acquisition.json

@@ -0,0 +1,330 @@
+{
+  "version": 1,
+  "phase": "complete",
+  "sourceCommit": "01ca48872fdceac9d7aa548ac12a5107173f2b31",
+  "host": "2.1.321 (Claude Code)",
+  "model": "claude-sonnet-5-5",
+  "seed": "portulan-m13-personas-20261110",
+  "prices": {
+    "uncached": 1,
+    "written5m": 1.25,
+    "written1h": 2,
+    "writtenUnknown": 1.25,
+    "read": 0.1,
+    "output": 5,
+    "coldFirstInput": 1.25
+  },
+  "routing": {
+    "model": "inherit",
+    "effort": null,
+    "omitClaudeMd": false,
+    "cacheLifetime": "5m"
+  },
+  "publication": {
+    "answerTransformation": "replace only known plugin-root and acquisition-root prefixes with <plugin> and <acquisition>",
+    "normalizedAnswers": [
+      "librarian-sealed-a"
+    ],
+    "note": "Original wording otherwise retained; original answers stay private with raw-answer digests. This presentation normalization changes no semantic artifact or score."
+  },
+  "qualityFrozenAt": "2026-11-10T07:29:01.199Z",
+  "continuationFrozenAt": "2026-11-10T07:49:23.195Z",
+  "order": [
+    "implementer-lower-bound-a",
+    "implementer-lower-bound-b",
+    "implementer-dedup-b",
+    "implementer-dedup-a",
+    "implementer-containment-a",
+    "implementer-containment-b",
+    "reviewer-verdict-b",
+    "reviewer-verdict-a",
+    "reviewer-clean-a",
+    "reviewer-clean-b",
+    "reviewer-boundaries-b",
+    "reviewer-boundaries-a",
+    "librarian-duplicate-a",
+    "librarian-duplicate-b",
+    "librarian-conflict-b",
+    "librarian-conflict-a",
+    "librarian-sealed-a",
+    "librarian-sealed-b"
+  ],
+  "taskSha256": {
+    "implementer-lower-bound": "4862de332848a8546671bbd33b6a2827d3bced336d65b92ed4aa8f02ce763290",
+    "implementer-dedup": "558e39b28c0f012f7fc5c0ceeec8898fd6fc77040305c587908b8795113afce8",
+    "implementer-containment": "2d991b5fc2fdb3e7d03dbebce8f9406d5ad1e55892d4dce1503a544b24cc525a",
+    "reviewer-verdict": "85e4e4f33f81550775e8b155906e2ecfe6c05661ce40da023e3723429a188102",
+    "reviewer-clean": "562e341017f1da94f7f8be4c6dd878db8c5ce1498596856c58f003c4256f28c2",
+    "reviewer-boundaries": "36d848b6d1d015225680efca9f459fb326b691906de006e429ce62f8f7bb793c",
+    "librarian-duplicate": "8332dd4efbcde4643088f07c07439d0abbdf7d703bf1dd4d332531196197512e",
+    "librarian-conflict": "4af2b7ce4790e069235825a0b5d64d61ec0a66f0e61294a313701d92d9eb6cca",
+    "librarian-sealed": "6408193ff1b6315658d34200116e15f7300e0a5c639fc36ae44391867a69828b"
+  },
+  "preflights": [
+    {
+      "key": "mechanism-medium",
+      "task": "mechanism",
+      "role": "implementer",
+      "effort": "medium",
+      "requestedModel": null,
+      "model": "claude-fable-5-2",
+      "parentEffort": "high",
+      "fresh": true,
+      "context": "context-bd202d2470debba47cf5529f97fd4759da5a81e041d200d112d8ade5c6dc2de0",
+      "childContext": "child-3e143bad7e549e19b35c2318b06a8121ad1695e3df8d41c473e09600cb9d2acf",
+      "state": "invalid",
+      "exit": 1,
+      "timedOut": false,
+      "started": "2026-11-10T07:15:36.623Z",
+      "ended": "2026-11-10T07:15:45.112Z",
+      "argv": [
+        "-p",
+        "This is a frozen evaluation. Invoke the Agent tool exactly once with these input values: {\"subagent_type\":\"portulan:implementer\",\"description\":\"Frozen persona artifact probe\",\"prompt\":\"Read <acquisition>/tasks/mechanism.json and follow its task exactly. This is an artifact proposal, not a completed repository change.\",\"run_in_background\":false,\"effort\":\"medium\"}. Do not supply a child model or change the effort. Wait for that foreground child to finish. Do not answer its task yourself or invoke another child. After completion reply only DONE.",
+        "--output-format",
+        "json",
+        "--effort",
+        "high",
+        "--plugin-dir",
+        "<plugin>",
+        "--tools",
+        "Agent,Read",
+        "--allowedTools",
+        "Agent(portulan:implementer),Read",
+        "--add-dir",
+        "<acquisition>/tasks",
+        "--max-budget-usd",
+        "1"
+      ],
+      "argvProvenance": "reconstructed-after-execution",
+      "instrumentSha256": "a02b6d2185b59e5b43a96a7cb3a5fcee7965c29a45b1533aa02c033369937f44",
+      "usage": {
+        "parent": [
+          {
+            "id": "request-832595efc5c05c77691a8e9c83ebd22210d0bf5aecccb43876f0e3860460dad8",
+            "model": "claude-fable-5-2",
+            "effort": "high",
+            "at": 1794294941008,
+            "uncached": 2,
+            "written5m": 83513,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 0,
+            "output": 245
+          }
+        ],
+        "child": [
+          {
+            "id": "request-4e55082b4084a7622c2d5623a1b91d9bb0ba9e4a6e8445799b10ca7305295b36",
+            "model": "claude-fable-5-2",
+            "effort": "medium",
+            "at": 1794294942508,
+            "uncached": 2,
+            "written5m": 13333,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 0,
+            "output": 314
+          },
+          {
+            "id": "request-9a157d22debc4d6232568c09309de1c0b38aeb1cfc73384bca3cbce633ad7eec",
+            "model": "claude-fable-5-2",
+            "effort": "medium",
+            "at": 1794294944421,
+            "uncached": 2,
+            "written5m": 5993,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 13333,
+            "output": 11
+          }
+        ]
+      },
+      "raw": {
+        "parent": "218d6fdf034153ffafe1b6ac5a96854180ea6a4310a4876cd888bcd6ce067d64",
+        "child": "324d3c6e00ba0ac72ed03478bffbb6dbf65e8173490467b460f242eaca0fd3ec",
+        "result": "40affabe76d7e056ab7867a593410a74ecb2e3eac8e725fcdad5aecf45f2021f"
+      },
+      "hostEstimateUsd": 1.0539046,
+      "reason": "error_max_budget_usd",
+      "markerSeen": true,
+      "sourceSnapshotTiming": "after-execution",
+      "freshContextEvidence": "one parent transcript and one child transcript; no resume option"
+    },
+    {
+      "key": "mechanism-medium",
+      "task": "mechanism",
+      "role": "implementer",
+      "effort": "medium",
+      "requestedModel": "claude-sonnet-5-5",
+      "model": "claude-sonnet-5-5",
+      "parentEffort": "high",
+      "fresh": true,
+      "context": "context-9521e523e92aab3fb2a046a983432e56511424e7a91df2c4a5a76143e4f87dff",
+      "childContext": "child-847477035c4dfacc71140c7d805005629bc874cf015a83acf28bd62a2190be0f",
+      "state": "complete",
+      "exit": 0,
+      "timedOut": false,
+      "started": "2026-11-10T07:22:15.599Z",
+      "ended": "2026-11-10T07:22:29.819Z",
+      "argv": [
+        "-p",
+        "This is a frozen evaluation. Invoke the Agent tool exactly once with these input values: {\"subagent_type\":\"portulan:implementer\",\"description\":\"Frozen persona artifact probe\",\"prompt\":\"Read <acquisition>/tasks/mechanism.json and follow its task exactly. This is an artifact proposal, not a completed repository change.\",\"run_in_background\":false,\"effort\":\"medium\"}. Do not supply a child model or change the effort. Wait for that foreground child to finish. Do not answer its task yourself or invoke another child. After completion reply only DONE.",
+        "--output-format",
+        "json",
+        "--session-id",
+        "<parent-context>",
+        "--effort",
+        "high",
+        "--plugin-dir",
+        "<plugin>",
+        "--tools",
+        "Agent,Read",
+        "--allowedTools",
+        "Agent(portulan:implementer),Read",
+        "--add-dir",
+        "<acquisition>/tasks",
+        "--max-budget-usd",
+        "1",
+        "--model",
+        "claude-sonnet-5-5"
+      ],
+      "argvProvenance": "recorded-before-execution",
+      "instrumentSha256": "fbde0bff490748c02803a6f953505eaa057031285a39957573785cd14aa388d8",
+      "usage": {
+        "parent": [
+          {
+            "id": "request-8d5746971407ca792e9da0fb449c6028995eee2f8d7b2ddda1f4e6da61937613",
+            "model": "claude-sonnet-5-5",
+            "effort": "high",
+            "at": 1794295341109,
+            "uncached": 2,
+            "written5m": 82083,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 0,
+            "output": 245
+          },
+          {
+            "id": "request-ac34361003d5e3e5e9d2ebe0ac70382e3e7aea7c418ec7c07c73ddc5813c5543",
+            "model": "claude-sonnet-5-5",
+            "effort": "high",
+            "at": 1794295346829,
+            "uncached": 2,
+            "written5m": 503,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 82083,
+            "output": 5
+          }
+        ],
+        "child": [
+          {
+            "id": "request-9c5e95a551188dc248414fc894df567dadbd76c16bf02cfaa91a1c3cfd946425",
+            "model": "claude-sonnet-5-5",
+            "effort": "medium",
+            "at": 1794295342891,
+            "uncached": 2,
+            "written5m": 13337,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 0,
+            "output": 314
+          },
+          {
+            "id": "request-f771db03c2ead365b17b2fcb5ef6f2e0d0d25b594161035e0a092fd115d2fc32",
+            "model": "claude-sonnet-5-5",
+            "effort": "medium",
+            "at": 1794295345372,
+            "uncached": 2,
+            "written5m": 5989,
+            "written1h": 0,
+            "writtenUnknown": 0,
+            "read": 13337,
+            "output": 11
+          }
+        ]
+      },
+      "raw": {
+        "result": "6b6faa6b1ee0157c9b931ab46e15e016df7f34b647bb34c1d7553ed6b2d81441",
+        "parent": "31fabf7e7f548051937823e01c49ec1b573d37d903a010eb72eecbdcde91b037",
+        "child": "f4af68b961cac7eacfec012332061b67b6ba2f34c2eeb7a9ffc9b43421b211a8"
+      },
+      "hostEstimateUsd": 0.27008800000000005,
+      "invocation": {
+        "type": "portulan:implementer",
+        "effort": "medium",
+        "promptSha256": "7bbcfe957dfaf278f89ae1cb9fb6902479eb1d8f53e131ce49d5f0aa74ff064d",
+        "model": null,
+        "background": false
+      },
+      "answer": "MECHANISM_OK",
+      "answerRawSha256": "72a7832e87334fc8c2dda2407f419a828f8fb30eda2eed583debfa49b4112c70",
+      "answerPresentationSha256": "72a7832e87334fc8c2dda2407f419a828f8fb30eda2eed583debfa49b4112c70",
+      "answerNormalized": false,
+      "transcriptDuplicates": {
+        "parent": 0,
+        "child": 2
+      },
+      "markerSeen": true
+    }
+  ],
+  "decodeRepair": {
+    "retainedKey": "implementer-lower-bound-a",
+    "approvedBy": "/root/checkpoint_520_open",
+    "approvalSha256": "9a8c8e99cfaec94ad350cac295d1f79dcca2396645c94578f575c9f69dab1b12",
+    "abortedJournalSha256": "6afb381b0e9d3eaefe2514e48d78295b01aec58ae2f878fdbb64c9fc667ad6ab",
+    "previousGrader": "9035deb344ea5a3ba5d31485836a5dda4d3dba6e06b20bfa78afb945ae1096b0",
+    "grader": "914d7744b73e9a79dd157038ed57c6f3034b2a05ce46fb4b5546a08b2e660978",
+    "observedOutcomes": 1,
+    "approval": "PASS",
+    "reruns": 0,
+    "semanticSuffixSha256": "3b8bde9accae1687d26b4621cafbadf7fb11e8ebd5dfd348ba3d15534f451146",
+    "explanation": "After one outcome, the original decoder rejected one JSON fence followed by prose. The same retained answer passes the revised decoder; all semantic predicates, cases and price code are byte-identical. The original false score remains. Independent approval covered exact decoder and continuation instrument hashes before the remaining seventeen calls."
+  },
+  "accounting": {
+    "calls": 20,
+    "parentContexts": 20,
+    "qualityChildren": 18,
+    "requests": 80,
+    "note": "Both probes, every parent and every child are included. Host USD values are lifecycle estimates, not subscription bills, and are not added to the declared-vector indexes."
+  },
+  "sources": {
+    "evals/ab/personas/corpus.json": "89554b37030657bd7167395e425461cb6f7acf5c223090c074ff054c7dac6b9b",
+    "evals/ab/personas/grader.mjs": "914d7744b73e9a79dd157038ed57c6f3034b2a05ce46fb4b5546a08b2e660978",
+    "core/personas/implementer.md": "bea6b78808fd6c390ebdfb53124d0b132e8a1454d78cecbabe04a632a0e2d45c",
+    "core/personas/reviewer.md": "6c364f370e61bc80a32697dc7d84eeb267ef3a2bbf48da16744356a90ce5078f",
+    "core/personas/librarian.md": "a79ef87153c2a5f52d36bd33b95fcdc2b0ed9bf79362a76bbf6a85459c094f46"
+  },
+  "bindings": {
+    "implementer": "3c3874f38defa6ea2bb42e45aa5a3b4267cc6794e1c7b11edc27440f861c0b97",
+    "reviewer": "436896a48ce036eeab69091fce7cee7482e8891efa2315da288378d9158f9a69",
+    "librarian": "7ad4e74c09482940444be0a1fe4de7eacc98e24aee145fdb6c25a162879053fc"
+  },
+  "instruments": {
+    "firstProbe": {
+      "sha256": "a02b6d2185b59e5b43a96a7cb3a5fcee7965c29a45b1533aa02c033369937f44",
+      "timing": "snapshotted-after-execution"
+    },
+    "secondProbe": {
+      "sha256": "fbde0bff490748c02803a6f953505eaa057031285a39957573785cd14aa388d8",
+      "timing": "frozen-before-execution"
+    },
+    "firstOutcome": {
+      "sha256": "2f0c0843c55ced112b8f5fc6872de6acb84f16319dcdd8edcf763bac4c37aa30",
+      "timing": "frozen-before-execution"
+    },
+    "continuation": {
+      "sha256": "34a45b79d5d3a762bbcd46222bdefe479cf031d1d0150665763b8de1c78981ab",
+      "timing": "approved-and-frozen-before-seventeen-untouched-cells"
+    },
+    "publisher": {
+      "sha256": "58ceb6aa7b361f627816fc92c1b7cdc755b3c473c5509b25c1a97c3c88a0d96c",
+      "timing": "after-acquisition; whitelisted normalization only"
+    }
+  },
+  "files": {
+    "implementer.json": "0b63097bec76852cccdd8edf3c7c9f20d83b9caae9536d1badb51a269003cdef",
+    "reviewer.json": "b9ea8b9d5b24ab33b67a1274b507b8416a482d349a9d64b4261b76ceac856774",
+    "librarian.json": "adf79f475a83dcc321dff1e4867d9af232052ffda6ffc92e5d2c8c75cf65a466"
+  }
+}

package/evals/ab/personas/bindings/implementer.txt

@@ -0,0 +1,43 @@
+---
+name: implementer
+description: Carries one task through the full loop — research, plan, implement, verify, learn — and hands back a verified change plus what it learned. Delegate a single unit of implementation work to it. It does not merge and it does not take gated actions.
+tools: Read, Glob, Grep, Edit, Write, Bash
+# compiled by `portulan compile` from core/personas/implementer.md; edit that file, then recompile
+model: "inherit"
+omitClaudeMd: false
+experimental:
+  cacheTtl: 5m
+---
+
+You are the **implementer** persona of the Portulan engine. Read
+`${CLAUDE_PLUGIN_ROOT}/core/personas/implementer.md` for your charter and
+`${CLAUDE_PLUGIN_ROOT}/core/operating/loop.md` for the loop you carry. This file binds that persona to
+this host; the doctrine lives there, and where the two ever disagree, the persona wins.
+
+Take **one task** from intent to a demonstrated change:
+
+1. **Research** before planning. Read the repository's workspace — its gate map, definition of done,
+   the repo card for this repository, and relevant memory — and never guess what you can check.
+2. **Plan** to a testable shape. If the acceptance criteria cannot be written as observable checks,
+   run the `clarify` skill rather than starting anyway.
+3. **Implement** at the altitude of the surrounding code. Writes are isolated: one branch, your own
+   files.
+4. **Verify** by climbing the hierarchy — it compiles, then tests pass, then the behaviour is
+   exercised. Run the workspace's verify recipe and read the output. Never report done on what you
+   could not walk a reviewer through line by line.
+5. **Learn**: write back what the task taught, as a draft. You draft memory and propose rule changes;
+   you never accept them. That gate is a human's.
+
+**What you do not do.** You do not merge your own work, and you do not take an action the workspace's
+gate map places in the Gated tier — merging, deleting a remote branch, publishing, changing settings,
+spending money, or anything that speaks for the team beyond this repository. You prepare such an action
+and hand it to the human. Which actions those are is the gate map's to say and not this file's, so read
+it rather than this list. Where the line falls — whether pushing a working branch is Auto, say, or the
+gate sits at the merge — is a property of the workspace you are standing in, not of this binding: these
+files ship with the plugin into repositories whose gate maps this one has never seen.
+
+**One honest limit of this binding.** The engine states your reach as capability classes; this host's
+allow-list is coarser. `Bash` is a single grant covering both running a verify recipe and running
+`gh pr merge`, so the tool list cannot express the Auto/Gated line by itself. The line is held by the
+gate map you read in step 1 and by the repository's platform floor — branch protection and required
+checks — not by this frontmatter. Treat the gate map as binding regardless of what this host permits.