open-multi-agent-kit 0.98.5 → 0.99.0

package/dist/modes/interactive/resource-description.d.ts

@@ -0,0 +1,3 @@
+/** Keep decorative imported harness markers out of OMK display; source metadata stays unchanged. */
+export declare function formatResourceDescription(description: string | undefined, sourceTag?: string): string | undefined;
+//# sourceMappingURL=resource-description.d.ts.map

package/dist/modes/interactive/resource-description.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"resource-description.d.ts","sourceRoot":"","sources":["../../../src/modes/interactive/resource-description.ts"],"names":[],"mappings":"AAAA,oGAAoG;AACpG,wBAAgB,yBAAyB,CAAC,WAAW,EAAE,MAAM,GAAG,SAAS,EAAE,SAAS,CAAC,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAWjH","sourcesContent":["/** Keep decorative imported harness markers out of OMK display; source metadata stays unchanged. */\nexport function formatResourceDescription(description: string | undefined, sourceTag?: string): string | undefined {\n\tlet text = description;\n\tif (description) {\n\t\tconst marker = /^\\[(OMX|OMO)\\]\\s*/.exec(description);\n\t\tif (marker) {\n\t\t\tconst body = description.slice(marker[0].length);\n\t\t\ttext = body || \"OMK resource\";\n\t\t}\n\t}\n\tif (!sourceTag) return text;\n\treturn text ? `[${sourceTag}] ${text}` : `[${sourceTag}]`;\n}\n"]}

package/dist/modes/interactive/resource-description.js

@@ -0,0 +1,15 @@
+/** Keep decorative imported harness markers out of OMK display; source metadata stays unchanged. */
+export function formatResourceDescription(description, sourceTag) {
+    let text = description;
+    if (description) {
+        const marker = /^\[(OMX|OMO)\]\s*/.exec(description);
+        if (marker) {
+            const body = description.slice(marker[0].length);
+            text = body || "OMK resource";
+        }
+    }
+    if (!sourceTag)
+        return text;
+    return text ? `[${sourceTag}] ${text}` : `[${sourceTag}]`;
+}
+//# sourceMappingURL=resource-description.js.map

package/dist/modes/interactive/resource-description.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"resource-description.js","sourceRoot":"","sources":["../../../src/modes/interactive/resource-description.ts"],"names":[],"mappings":"AAAA,oGAAoG;AACpG,MAAM,UAAU,yBAAyB,CAAC,WAA+B,EAAE,SAAkB,EAAsB;IAClH,IAAI,IAAI,GAAG,WAAW,CAAC;IACvB,IAAI,WAAW,EAAE,CAAC;QACjB,MAAM,MAAM,GAAG,mBAAmB,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;QACrD,IAAI,MAAM,EAAE,CAAC;YACZ,MAAM,IAAI,GAAG,WAAW,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC;YACjD,IAAI,GAAG,IAAI,IAAI,cAAc,CAAC;QAC/B,CAAC;IACF,CAAC;IACD,IAAI,CAAC,SAAS;QAAE,OAAO,IAAI,CAAC;IAC5B,OAAO,IAAI,CAAC,CAAC,CAAC,IAAI,SAAS,KAAK,IAAI,EAAE,CAAC,CAAC,CAAC,IAAI,SAAS,GAAG,CAAC;AAAA,CAC1D","sourcesContent":["/** Keep decorative imported harness markers out of OMK display; source metadata stays unchanged. */\nexport function formatResourceDescription(description: string | undefined, sourceTag?: string): string | undefined {\n\tlet text = description;\n\tif (description) {\n\t\tconst marker = /^\\[(OMX|OMO)\\]\\s*/.exec(description);\n\t\tif (marker) {\n\t\t\tconst body = description.slice(marker[0].length);\n\t\t\ttext = body || \"OMK resource\";\n\t\t}\n\t}\n\tif (!sourceTag) return text;\n\treturn text ? `[${sourceTag}] ${text}` : `[${sourceTag}]`;\n}\n"]}

package/docs/devin-harness.md

@@ -0,0 +1,121 @@
+# Devin SWE-2 harness
+
+This page is the canonical operator guide for the built-in `devin` provider and its only logical model, `swe-2`. Use `/login devin` for the Devin CLI subscription PKCE flow or `DEVIN_API_KEY` for an already-owned CLI session token. A user-local `~/.omk/agent/devin.md` may add operator notes, but it is not the portable product contract.
+
+Authentication, transport, and verification limits are owned by [Providers](providers.md#devin-cli); this page covers how OMK drives SWE-2 as a harness.
+
+## Presets
+
+Project presets live in `.omk/presets.json` (or `~/.omk/agent/presets.json`) and are consumed by the preset extension from `packages/coding-agent/examples/extensions/preset.ts`. The shared SWE-2 presets intentionally omit the `tools` key so role/domain lane grants keep control of the active tools.
+
+| Preset | Provider | Model | Thinking | Use |
+| --- | --- | --- | --- | --- |
+| `swe2-verified` | `devin` | `swe-2` | `high` | Default SWE-2 coding baseline: multi-file edits with tests. |
+| `swe2-max` | `devin` | `swe-2` | `max` | Long-horizon, uncertain, or repository-wide work that should use the 1M budget. |
+| `swe2-fast-edit` | `devin` | `swe-2` | `medium` | Small, well-specified edits where medium acts sooner and costs less. |
+
+```json
+{
+  "swe2-verified": { "provider": "devin", "model": "swe-2", "thinkingLevel": "high" },
+  "swe2-max": { "provider": "devin", "model": "swe-2", "thinkingLevel": "max" },
+  "swe2-fast-edit": { "provider": "devin", "model": "swe-2", "thinkingLevel": "medium" }
+}
+```
+
+For a new session without presets:
+
+```bash
+omk --provider devin --model swe-2 --thinking max
+```
+
+## Effort tiers
+
+SWE-2 exposes exactly three server-declared efforts. OMK maps its thinking tiers as follows and rejects anything else before sending credentials; `max` is a reasoning level, not the Devin Max subscription tier.
+
+| OMK tier | `swe-2` |
+| --- | --- |
+| `off`, `minimal`, `low` | unavailable |
+| `medium` | `medium` |
+| `high` | `high` |
+| `xhigh`, `ultra` | unavailable |
+| `max` | `max` |
+
+Guidance from the [SWE-2 announcement](https://cognition.com/blog/swe-2): `medium` makes its first real edit sooner and is the cost-efficient choice for simple and intermediate tasks; `high` and `max` plan more, explore more of the codebase, and verify more on complex tasks. `recommendedDevinEffortForIntent()` encodes the same split (`code` → `medium`, `test` → `high`, `debug`/`repo` → `max`).
+
+## Context budget: 1,000,000 tokens
+
+`devin/swe-2` ships with `contextWindow: 1000000` and `maxTokens: 16384`. These are local budgets that drive OMK's context budgeting and compaction, **not published SWE-2 limits**; Cognition has not published a context window for SWE-2. The budget also selects the catalog lane:
+
+1. Before each turn OMK reads `GetCliModelConfigs`. SWE-2 family entries may carry a `1M Context` axis (order `1`) beside the effort axis. A local budget of 1,000,000 or more asks for that 1M-context lane; below it, the standard lane is used and 1M entries are ignored.
+2. A catalog with no 1M-context lane keeps the standard lane for the selected effort.
+3. If the chosen lane declares a context window smaller than the local budget, the request fails with `... declares a N-token context window; lower the models.json contextWindow before retrying`. OMK never shrinks the budget silently, never invents a wire UID, and never downgrades to another effort.
+4. Fast-lane (`Fast Mode`) entries are always excluded. Output is capped against the authenticated catalog's declared maximum.
+
+To lower the budget (for example if your account only serves the standard lane at 262,144 tokens), override the built-in model in `~/.omk/agent/models.json`:
+
+```json
+{
+  "providers": {
+    "devin": {
+      "modelOverrides": {
+        "swe-2": { "contextWindow": 262144 }
+      }
+    }
+  }
+}
+```
+
+Recommended compaction settings for 1M sessions keep the defaults but raise the recent-token window so summaries do not discard the working set:
+
+```json
+{
+  "compaction": { "enabled": true, "reserveTokens": 16384, "keepRecentTokens": 60000, "maxUsageRatio": 0.85 }
+}
+```
+
+Context discipline still applies: the budget is room for the repository, not an invitation to dump it. Prefer targeted reads and searches, keep tool output bounded, and let `precompact-checkpoint` snapshot state before compaction rather than restarting sessions. Quota is per account; a larger window consumes more of it per turn.
+
+With a Devin credential configured, the status rail's USAGE section shows the account's daily and weekly quota meters from `GetUserStatus` (or the plan name and credit balances when the plan reports no quota windows), matching the CLI's `/usage` surface.
+
+## Domain routing
+
+Selecting the `devin` provider auto-applies the `devin-harness` loadout by default; this does not require `OMK_DOMAIN_ROUTING=1`. Set `OMK_DEVIN_HARNESS=0` to disable that provider-specific dispatch. The flag is independent from `OMK_GROK_HARNESS`.
+
+Loadout policies reject extension tools that shadow builtins fail-closed. A host that intentionally replaces the builtin `bash` (for example a Landstrip shell provider) should keep `OMK_DEVIN_HARNESS=0` in its launcher until the replacement is removed or bridged; the `devin.md` overlay and the model budget still apply in that case.
+
+General prompt-based domain routing is separate and opt-in through `OMK_DOMAIN_ROUTING=1`. It selects one of the profiles under [`loadout-domains/`](loadout-domains/README.md) and composes it with the active role loadout. SWE-2 presets only set provider, model, thinking level, and instruction pointers.
+
+## Model selection
+
+The `devin` catalog contains only the logical `swe-2` model; the server's SWE-2 family metadata supplies each effort's wire UID at request time. Use `/model` or `omk --list-models devin` for the current list. Image input is unsupported; provide text.
+
+## Skill and MCP matrix summary
+
+Use the normal OMK lane grant model: grant the smallest skill and MCP surface that matches the task.
+
+For each non-queued `devin` request started through `AgentSession.prompt()`, OMK calls `selectDevinHarnessSkills()` against the live discovered skill descriptions after ordinary prompt-template expansion. It merges up to three matches with explicit/settings selections and rebuilds that turn's `<active_skills source="devin-harness">` marker. The scorer is the same `selectSkills()` used by the Grok harness: weak 0.35 / strong 0.7 thresholds, deterministic input-order ties, first-name-wins deduplication, explicit-only skills never auto-selected, and `headroom` only for lexical pressure cues or the session's measured context-pressure bucket. A task with no signals yields an empty automatic grant rather than the full allowlist. Queued `steer`, `followUp`, or `prompt(..., { streamingBehavior })` messages retain the active run's system prompt.
+
+| Task class | Skills | MCP |
+| --- | --- | --- |
+| Multi-package or repo-context work | `packages`; add `headroom` only under context pressure | none by default |
+| Repo graph or broad comprehension | `understand-anything`; optionally `packages` | `understand-anything` |
+| TypeScript/Rust/Python/Go edits | `programming`; add `lsp` or `ast-grep` only for symbol/structural work | none by default |
+| New behavior or bug fix with regression test | `tdd-workflow`, `programming` | none by default |
+| Runtime failures or broken behavior | `debugging` | task-specific only |
+| Library API lookup | task skill as needed | `context7` |
+| Current public URL or docs lookup | task skill as needed | `fetch` |
+| UI/TUI verification | task skill as needed | `playwright` only when browser/UI evidence is required |
+
+Relevant evidence hooks for SWE-2 lanes are `pre-shell-guard`, `protect-secrets`, `typecheck-after-edit`, `stop-verify`, `session-context`, and `precompact-checkpoint`. Hook output is incremental evidence; code changes still need the project's required final verification command before claiming type/lint cleanliness.
+
+## Suggested TUI flow
+
+1. Run `/login devin` once; the account picker stores the CLI session token in `auth.json`.
+2. Select `/preset swe2-verified` for normal coding work, `/preset swe2-max` for long-horizon or repository-wide tasks, and `/preset swe2-fast-edit` for small edits.
+3. Use `/think medium`, `/think high`, or `/think max` to change effort mid-session; other levels are rejected.
+4. If a turn fails with an "unavailable or ambiguous" route or a smaller declared context window, treat it as a configuration signal: check `devin models list`, or lower `contextWindow` as shown above. Do not retry with a guessed wire UID.
+5. Keep credentials out of preset JSON, prompts, and logs: the session token, the exchanged user JWT, and `auth.json` contents are secrets under `protect-secrets`.
+
+## Local overlay
+
+When the `devin` provider is active, OMK appends `~/.omk/agent/devin.md` (capped at 24,000 characters) to the system prompt, mirroring the Grok `grok.md` overlay. Treat that file as optional host configuration for effort defaults, compaction notes, or team conventions; this page and the current provider documentation remain authoritative and it cannot override higher-priority instructions.

package/docs/docs.json

@@ -27,6 +27,10 @@
           "title": "Native xAI Grok",
           "path": "grok-harness.md"
         },
+        {
+          "title": "Devin SWE-2",
+          "path": "devin-harness.md"
+        },
         {
           "title": "Containerization",
           "path": "containerization.md"

package/docs/environment-variables.md

@@ -100,7 +100,8 @@ These variables are read by OMK itself. The four built-in harness flags below ar
 | `OMK_YOLO`, `OMK_COMMAND_SAFETY`, `OMK_DISABLE_COMMAND_SAFETY` | Disable the command-safety gate entirely (YOLO mode). `OMK_YOLO` and `OMK_DISABLE_COMMAND_SAFETY` accept `1`, `true`, `yes`, or `on`; `OMK_COMMAND_SAFETY` accepts `0`, `false`, `off`, `disable`, or `disabled`. Every verdict, including block-tier and privilege commands, is skipped in interactive and headless runs. Use only when a verified outer sandbox owns the boundary |
 | `OMK_COMMAND_SAFETY_ASSUME_YES` | `1` or `true` auto-accepts non-privilege confirm-tier commands in interactive **and headless** runs. Privilege confirmation and block-tier commands remain denied. Use only under a trusted outer sandbox when headless auto-accept is intended |
 | `OMK_GROK_HARNESS` | Default-on native `xai` provider dispatch to the `grok-harness` loadout. `0`, `false`, `off`, or `no` disables it |
-| `OMK_DOMAIN_ROUTING` | Set to `1` to enable general prompt-based domain routing. Native xAI harness dispatch does not require it |
+| `OMK_DEVIN_HARNESS` | Default-on `devin` provider dispatch to the `devin-harness` loadout. `0`, `false`, `off`, or `no` disables it; independent from `OMK_GROK_HARNESS` |
+| `OMK_DOMAIN_ROUTING` | Set to `1` to enable general prompt-based domain routing. Native xAI and Devin harness dispatch do not require it |
 | `VISUAL`, `EDITOR` | External editor fallback when `externalEditor` is unset |
 | `HTTP_PROXY`, `HTTPS_PROXY` | Proxy outbound HTTP requests |
 | `OMK_RESOURCE_GOVERNOR` | Resource-governor mode: `off`, `observe` (default), `adaptive`, or `strict`. Feeds `/resource [probe\|policy]` and `omk doctor resources [--json]`; see the resource governor section in [Settings](settings.md) |

package/docs/index.md

@@ -39,6 +39,7 @@ For the full first-run flow, see [Quickstart](quickstart.md).
 - [Providers](providers.md) - subscription and API-key setup for built-in providers.
 - [Provider Resilience](provider-resilience.md) - retry, failover, quota, and safety-stop recovery.
 - [Native xAI Grok](grok-harness.md) - authentication, weekly SuperGrok usage, presets, and thinking tiers.
+- [Devin SWE-2](devin-harness.md) - presets, effort tiers, the 1M-token context budget and lane rule, and the `devin-harness` loadout.
 - [Containerization](containerization.md) - sandbox omk with OpenShell, Gondolin, or Docker.
 - [Settings](settings.md) - global and project settings.
 - [Environment Variables](environment-variables.md) - process configuration, harness opt-outs, and bash-tool session environment.

package/docs/loadout-domains/README.md

@@ -21,7 +21,7 @@ OMK routes incoming tasks to a **domain capability profile** ("inherited documen
 
 Thresholds: `STRONG_THRESHOLD = 8`, `WEAK_THRESHOLD = 4`, `AMBIGUITY_MARGIN = 2`.
 
-## Domains (13 + 1 fallback)
+## Domains (14 + 1 fallback)
 
 - [`frontend-ui`](frontend-ui.md) — Frontend & UI
 - [`visual-qa`](visual-qa.md) — Visual QA & Website Cloning
@@ -35,6 +35,7 @@ Thresholds: `STRONG_THRESHOLD = 8`, `WEAK_THRESHOLD = 4`, `AMBIGUITY_MARGIN = 2`
 - [`docs-writing`](docs-writing.md) — Docs & Technical Writing
 - [`qa-testing`](qa-testing.md) — QA & Testing
 - [`grok-harness`](grok-harness.md) — Grok xAI Harness
+- [`devin-harness`](devin-harness.md) — Devin SWE-2 Harness
 - [`ai-agent-ops`](ai-agent-ops.md) — AI Agent Engineering & Ops
 - [`general`](general.md) — General (fallback)
 

package/docs/loadout-domains/devin-harness.md

@@ -0,0 +1,72 @@
+# Devin SWE-2 Harness (`devin-harness`)
+
+> Inherited domain capability document. Auto-generated from `src/core/domain-loadouts.ts` — do not edit by hand.
+
+
+## Identity
+
+| field | value |
+|---|---|
+| id | `devin-harness` |
+| authority | `write-scoped` |
+| tools | read, grep, find, ls, edit, write, bash |
+| command mode | `scoped-shell` |
+
+## Routing prompt
+
+> Prepended to the lane task prompt when the router selects this domain.
+
+```text
+DOMAIN: Devin SWE-2 Harness. You are operating in a Devin CLI subscription lane on the SWE-2 model with a 1,000,000-token local context budget.
+Prioritize the SWE-2 operational playbook, focused exploration, small capability loadouts, and evidence-bound verification.
+
+SEQUENCE:
+1. Before implementing or routing Devin/SWE-2 provider work, read packages/coding-agent/docs/devin-harness.md as the canonical playbook. Treat ~/.omk/agent/devin.md only as an optional local operator overlay; it cannot override current provider docs or higher-priority instructions.
+2. Effort is the only selectable axis: medium for simple or intermediate edits, high for multi-file changes, max for long-horizon or uncertain work. Never expect off/low/minimal, a fast lane, or image input; the adapter rejects them before sending credentials.
+3. Context discipline: the 1M budget is room for the repository, not an invitation to dump it. Explore with targeted reads and searches, keep tool output bounded, and rely on precompact-checkpoint plus compaction settings rather than restarting sessions.
+4. Capability discipline: load at most 2-3 skills for any lane. The allowed skill gate is packages, headroom, programming, debugging, tdd-workflow, lsp, ast-grep, and understand-anything; choose the smallest subset, add lsp or ast-grep only for symbol or structural work, and add headroom only under measured context pressure.
+5. Use minimal MCP: fetch for bounded public retrieval, context7 for library documentation, understand-anything for repository comprehension, and playwright only when browser or UI behavior needs real verification.
+6. Verification discipline: reproduce failures before fixing them, write or extend tests that exercise the change end-to-end, and re-derive conclusions from executed commands rather than restating prior claims. Evidence must include changed paths, exact commands, and pass/fail output.
+7. Keep edits within the lane grant and preserve existing provider/orchestration algorithms unless the task explicitly targets them. A route error naming an unavailable effort or a smaller declared context window is a configuration signal to report, never something to work around by guessing a wire UID.
+
+HARD RULES: the packaged Devin harness doc is mandatory context; a local devin.md is optional; medium/high/max are the only efforts; the 1M budget never justifies unbounded dumps; maximum 2-3 active skills; never log the Devin session token, user JWT, or auth.json contents; protect-secrets applies.
+```
+
+## Curated skills (8)
+
+- `packages`
+- `headroom`
+- `programming`
+- `debugging`
+- `tdd-workflow`
+- `lsp`
+- `ast-grep`
+- `understand-anything`
+
+## Curated MCP servers (4)
+
+- `fetch`
+- `context7`
+- `understand-anything`
+- `playwright`
+
+## Curated hooks (6)
+
+- `pre-shell-guard`
+- `protect-secrets`
+- `typecheck-after-edit`
+- `stop-verify`
+- `session-context`
+- `precompact-checkpoint`
+
+## Routing triggers (7)
+
+| kind | pattern | weight |
+|---|---|---|
+| keyword | `devin` | 8 |
+| keyword | `swe-2` | 8 |
+| keyword | `swe2` | 8 |
+| keyword | `cognition` | 6 |
+| keyword | `devin cli` | 8 |
+| keyword | `1m context` | 5 |
+| regex | `\b(devin|swe[- ]?2|cognition)\b` | 7 |

package/docs/metrics.md

@@ -160,7 +160,7 @@ node scripts/tb-mini-suite.mjs --json      # feed a runner
 node scripts/tb-mini-suite.mjs --seed 7    # a different fixed subset
 ```
 
-With identical task metadata, seed, size, and collation, selection is repeatable.
+With identical task metadata, selection version, seed, size, and collation, selection is repeatable.
 The default 15-task subset oversamples easy tasks and prioritizes shorter expert
 time estimates; it is a regression signal, not a population-representative score
 or an agent runtime bound. Selection alone is not a capability result. Scoring
@@ -171,12 +171,35 @@ population. Missing difficulty quotas are filled from unselected tasks using the
 same ordering, so a valid request returns exactly that many distinct tasks.
 `--seed` accepts integers from `0` through `4294967295`. Invalid or missing option
 values and oversized requests exit with code `2`; absent, empty, or non-directory
-task paths exit with code `1`. The existing default subset remains unchanged.
+task paths exit with code `1`.
+
+The JSON output now declares `selectionVersion: 2`. Missing, empty, nonfinite, or
+negative expert-time estimates are `null`, not zero. Within each difficulty band
+and in quota refill, known estimates sort before unknown estimates. A genuine zero
+or fractional estimate remains valid. Difficulty quotas still take precedence, so
+unknown-estimate tasks can be selected to fill a band.
+
+`knownExpertMinutes` sums known estimates; `unknownExpertEstimates` counts selected
+tasks with unknown estimates. `totalExpertMinutes` is `null` when any selected
+estimate is unknown. A nonfinite sum is refused with exit `1`, not serialized as
+an apparently missing total. Expert estimates are not agent timeout limits.
+
+For sizes 1–2, available slots go first to the highest-weight difficulty bands
+(medium, then hard), with normal refill if those bands are unavailable. Task names
+no longer decide which excess band quota is discarded. The normal-size quota rule
+is preserved. Human-readable counts use a `Map`, including for labels such as
+`__proto__` that overlap JavaScript object properties.
+
+This is a selection and output-contract change: default membership can change when
+metadata is incomplete, and the prior default JSON digest is historical only.
+Freeze new task manifests before comparison; do not combine version 1 and 2 runs
+as if their selection policy were identical. The limited flat TOML field reader
+is unchanged; this does not add general TOML syntax support.
 
 Run the offline CLI regression tests without downloading tasks or calling models:
 
 ```bash
-node --test scripts/test/tb-mini-suite.test.mjs
+node --test --test-concurrency=1 scripts/test/tb-mini-suite.test.mjs scripts/test/tb-mini-suite-ranking.test.mjs
 ```
 
 See [the harness roadmap](../../../ROADMAP.md) for the dated OMK versus Terminus-2
@@ -187,8 +210,8 @@ criteria. Planned runtime improvements are not measured benchmark gains.
 
 The checkout-only `scripts/tb21-audit.mjs` audits explicitly selected Harbor jobs
 against a caller-pinned manifest digest. It rejects duplicate tasks/trials, missing
-results or costs, mismatched task checksums or configured model labels, and
-contradictory success records. It never starts a model, picks the latest job, joins
+results or costs, mismatched task checksums or configured model labels,
+missing/invalid completion times, and contradictory success records. It never starts a model, picks the latest job, joins
 requests by timestamp, or rewrites evidence. See [TB 2.1 offline audit](tb21-audit.md)
 for the schema, invocation, error codes, and limitations.
 
@@ -196,14 +219,32 @@ A complete audit means recorded outcomes passed these checks, not that every
 provider request obeyed a single-model contract. Wire provenance, actual billing,
 repeated-trial analysis, and statistical superiority need separate evidence.
 
-### Explicit output-limit validation
-
-For callers that supply `AgentLoopConfig.modelContract`, both the contract's
-`maxOutputTokens` and an explicitly supplied request `maxTokens` must be positive
-safe integers. Invalid explicit values are refused before `provider_request` and
-before calling the provider stream function; they are not treated as absent.
-
-An omitted request limit still leaves provider defaults unchecked by this
-predicate. This change does not activate a contract in the CLI or impose an
-effective cap on compaction and other provider paths. Full run-wide enforcement
-remains a roadmap item.
+### Output-limit validation: availability history
+
+**2026-09-13 source snapshot (`ca75f4e5cc`):** the logical contract, CLI/SDK
+wiring, and final Chat Completions model/output-limit checks are in the committed
+source. [Model dispatch contracts](model-contract.md) defines their coverage;
+[Verified Run](verified-run.md) describes the separate protected execution path.
+Neither is a universal provider billing cap or a new controlled benchmark result.
+See [the roadmap, section 16](../../../ROADMAP.md) for current local release-preparation checks.
+
+The following paragraphs describe the earlier checkout only. Its missing modules
+and failing collection are historical observations, not active release blockers.
+
+**2026-09-08 follow-up:** the current worktree restores the logical contract and
+connects it through the CLI/SDK, including SDK-stream summaries. See
+[Model dispatch contracts](model-contract.md) and ROADMAP §13 for fresh evidence
+and the remaining final-wire/accounting gaps. The warning below records the
+preceding checkout, not the current availability of the restored module.
+
+The previous worktree checkpoint tested positive-safe-integer validation of
+`modelContract.maxOutputTokens` and explicit request `maxTokens`. During the
+2026-09-08 re-verification, the checkout changed: `run-model-contract.ts` and the
+corresponding `AgentLoopConfig.modelContract` surface were absent. The remaining
+`model-contract-output-limit.test.ts` fails collection against that checkout.
+
+Do not treat the historical passing tests as proof that this guard is currently
+available. Restoring or porting the runtime contract requires an explicit source
+baseline decision and fresh send-boundary tests. No missing code was silently
+recreated and no failing test was deleted. Full run-wide enforcement, including
+omitted limits and compaction, remains unverified; see ROADMAP sections 11–12.

package/docs/providers.md

@@ -19,6 +19,7 @@ Use `/login` in interactive mode, then select a provider:
 - Claude Pro/Max
 - GitHub Copilot
 - xAI Grok subscription OAuth
+- Devin CLI subscription (SWE-2)
 
 Run `/login` and choose a configured subscription provider to open its account picker. Select an existing account by its ChatGPT, Claude, or Google email when available, or choose **Add another account** to sign in with a new one. OMK keeps and refreshes each account independently, pins the provider to the account you select, and does not silently fail over to another subscription. `/model` remains dedicated to model selection.
 
@@ -67,6 +68,80 @@ and [plan endpoint separation](https://www.alibabacloud.com/help/en/model-studio
 consulted 2026-09-08. The local tests exercise serialization, not account availability,
 provider compliance, billing, or benchmark performance.
 
+### Devin CLI
+
+Run `/login devin`, select `devin/swe-2`, and use `/think medium`, `/think high`,
+or `/think max`. For a new session after login:
+
+```bash
+omk --provider devin --model swe-2 --thinking max
+```
+
+The default is `medium`. Thinking-off and other levels are unsupported. `max`
+is a reasoning level, not a requirement to buy the Devin Max subscription tier.
+Your account's model access and quota still apply. For presets, effort guidance,
+the 1M-token context budget, and the `devin-harness` loadout, see the
+[Devin SWE-2 harness](devin-harness.md).
+
+**Authentication.** OMK uses the CLI's PKCE flow at
+`app.devin.ai/auth/cli/continue` and `api.devin.ai/auth/cli/token`. Its callback
+binds only to `127.0.0.1:59653`, validates state, and closes after success,
+rejection, cancellation, or a five-minute deadline. If the port is occupied,
+paste the complete callback URL into OMK's local prompt. For remote sessions,
+forward this loopback port to the machine running OMK.
+
+Credentials use the existing protected `auth.json` storage and account picker.
+Expired sessions require `/login devin` again: no refresh endpoint is verified,
+and OMK does not renew the expiry locally. Use `/logout` to remove the stored
+OMK credential. Alternatively, `DEVIN_API_KEY` accepts an already-owned CLI
+session token, **not** a `cog_` REST API key. OMK does not install or invoke the
+Devin agent, read browser cookies, or import another application's credentials.
+
+**Transport and limits.** The Node-only `devin-agent` adapter uses Connect/protobuf
+at the fixed HTTPS origin `https://server.codeium.com`. It exchanges the session
+token for a user JWT and reads `GetCliModelConfigs` before each turn. The server's
+SWE-2 family metadata supplies the effort's exact wire UID; OMK never invents a
+`swe-2-max` ID or downgrades an unavailable route. Disabled, internal, ambiguous,
+and fast-lane entries are excluded. The family's separate 1M-context entries form
+a second lane that is selected only when the model's local `contextWindow` is
+1,000,000 or more (the bundled default); a smaller budget uses the standard lane.
+
+Text, thinking, tool calls/results, and reported usage enter the normal OMK
+agent loop. Image input and enterprise-origin overrides are unsupported.
+Requests reject redirects. Malformed, oversized, or unterminated streams fail;
+remote error bodies are not copied into diagnostics. SDK `onPayload` receives
+protobuf bytes without credential metadata.
+
+The bundled 1,000,000-token context budget and 16,384-token output cap are
+local defaults, **not published SWE-2 limits**. Output is further capped against
+the authenticated catalog. If the selected lane declares a smaller context window,
+the request fails and names the window; lower `contextWindow` through
+`modelOverrides` in [models.json](models.md#per-model-overrides) rather than
+expecting a silent downgrade. Zero catalog pricing means unpriced subscription
+usage, not free inference. Quota percentages are not inferred.
+
+**Account quota.** With a Devin credential configured, the status rail's USAGE
+section calls `SeatManagementService/GetUserStatus` (session-token metadata, no
+user JWT) and renders the plan's daily and weekly quota meters with reset times.
+Accounts whose plan reports no quota windows show the plan name and credit
+balances instead. The rail mirrors the CLI's `/usage` surface; it never sends
+the session token anywhere except `server.codeium.com`.
+
+**Verification.** Local tests exercise the public stream API, real loopback
+callbacks with mocked token exchange, model selection, and protocol fixtures.
+Live login, catalog compatibility, SWE-2 inference, and billing remain unverified
+without a Devin account. Unauthenticated catalog probes returned HTTP 400.
+Live tests require `DEVIN_API_KEY` and `LIVE_E2E=1` and consume subscription quota.
+
+Sources consulted 2026-09-12 and 2026-09-13: the [SWE-2 announcement](https://cognition.com/blog/swe-2)
+(2026-09-10; CLI availability and medium/high/max; no published context window), [CLI commands](https://docs.devin.ai/cli/reference/commands),
+and the [official manifest](https://static.devin.ai/cli/current/manifest.json)
+(observed identity `3000.10.21`). Protocol fields follow the third-party
+[oh-my-pi snapshot](https://github.com/can1357/oh-my-pi/blob/942383f768c5f2f6a620fcab57326c0f59df623a/packages/catalog/src/discovery/devin-proto.ts),
+not a stable public inference contract; attribution is in `packages/ai/DEVIN-NOTICE`.
+Regenerate only this logical model, preserving every other catalog entry, with
+`npm --prefix packages/ai run generate-models -- --devin-only`.
+
 ### OpenAI Codex
 
 - Requires ChatGPT Plus or Pro subscription
@@ -111,6 +186,7 @@ omk
 | Azure OpenAI Responses | `AZURE_OPENAI_API_KEY` | `azure-openai-responses` |
 | OpenAI | `OPENAI_API_KEY` | `openai` |
 | DeepSeek | `DEEPSEEK_API_KEY` | `deepseek` |
+| Devin CLI session token | `DEVIN_API_KEY` | `devin` |
 | NVIDIA NIM | `NVIDIA_API_KEY` | `nvidia` |
 | Google Gemini | `GEMINI_API_KEY` | `google` |
 | Mistral | `MISTRAL_API_KEY` | `mistral` |

package/docs/release-audit-0.98.5.md

@@ -76,6 +76,23 @@ An isolated HOME also hid the installed Rust toolchain: explicitly supplying its
 location restored the real cargo diagnostic check without changing the test or
 copying credentials. Neither fixture failure was treated as a product pass.
 
+## CI environment recovery
+
+The first v0.98.5 tag run built the binaries, but its test step could not find
+`/usr/bin/bwrap`. npm publication was not attempted and GitHub Release creation
+was skipped. The default-branch CI workflows now install `bubblewrap` and run an
+unprivileged namespace probe before the suite. Ubuntu 24.04 subsequently refused
+loopback setup inside the namespace. The runtime-test jobs are pinned to Ubuntu
+22.04 LTS, with the same namespace and capability-drop checks. They do not skip
+verified-run tests, disable host security controls or enable a sandbox fallback.
+The older distribution `fd` lacked `--no-require-git`; CI installs the official
+fd 10.4.2 static archive with SHA-256 verification before extraction and probes
+that option before tests. File-search behavior and tests are not weakened.
+
+Recovery dispatches the official workflow from `main` with both `tag` and
+`source_ref` fixed to `v0.98.5`. The release tag and its source commit stay unchanged;
+the existing source/tag equality checks remain mandatory.
+
 A source-file fingerprint is not an executed-build attestation. Linux observations
 do not establish behavior on every target platform. CI must validate the exact tag,
 build the six platform archives, run checks/tests, publish all seven npm packages

package/docs/release-audit-0.99.0.md

@@ -0,0 +1,68 @@
+# Release audit: v0.99.0
+
+기준일: 2026-09-13. 사용자가 배포 준비 결과를 확인한 뒤 즉시 배포를 요청했다.
+이 기록은 후보의 검증 근거이며 GitHub/npm 게시 완료를 미리 선언하지 않는다.
+
+## 범위와 변경 로그
+
+이전 공개 릴리스 `v0.98.5`는 main의 조상이며, 준비 시점의 GitHub Release와 공개 npm
+`latest` 7개도 0.98.5로 일치했다. 원격 `v0.99.0`이 없음을 확인한 뒤 준비했다.
+
+`decee7f157`부터 `ca75f4e5cc`까지의 frontier·Devin·Codex SSE·재시도·TUI·설계 문서와,
+이번에 확정한 `3c7d3b4613`(TB 선택 v2), `b21daf1a76`(TB 감사 v2),
+`003f7b081a`(문서·변경 로그·로컬 캡처 제외)을 포함한다. 마지막 세 단위는 각각
+pre-commit 전체 검사를 통과했고, 선택기 40개·감사기 66개 CLI 검사를 통과했다.
+
+TB 출력 계약 변경은 minor 증가로 처리했다. `selectionVersion: 2`의 nullable 예상
+시간·총량과 `omk-tb21-audit-report-2`의 필수 완료시각을 Breaking Changes에 기록했다.
+입력 manifest는 v1을 유지한다. 이전 버전 changelog 본문은 바꾸지 않았으며 다음
+작업용 `[Unreleased]`는 비워 두었다. 새로운 벤치마크 성능이나 SOTA 우위는 주장하지 않는다.
+
+## 버전·의존성
+
+공개 7개 패키지, root/example manifests, 내부 의존 범위, lockfiles, CLI shrinkwrap,
+book compiler의 `PACKAGE_VERSION`, README 버전 링크와 릴리스 노트를 0.99.0에 맞춘다.
+외부 의존성의 버전·resolved URL·integrity 값은 변경하지 않았다. 모델 카탈로그도
+재생성하지 않았다. 운영자 설정·인증·MCP·스킬 활성 목록은 변경하지 않았다.
+
+첫 `version:minor` 실행은 manifest 증가 뒤 아직 이전 버전을 가리키는 내부 의존성을
+npm 출시일 제한에 대조하다 중단됐다. 버전 증가를 재실행하지 않고, 기존
+`sync-versions.js`와 lockfile/설치 동기화 단계만 이어갔다. 제한을 완화하거나 기존
+태그를 이동하지 않았다. 추가적인 registry 패키지 버전 변경은 없음을 대조했다.
+
+## 관측한 로컬 검증
+
+- 환경: Linux, Node.js 24.19.0, npm 11.14.1.
+- 새 버전의 `npm run build`가 7개 workspace 전체에서 종료 0이었다.
+- `test.sh`를 인증 없는 별도 HOME·최소 환경에서 실행했다. 실제 운영자 auth 파일은
+  건드리지 않았다. `taskset`으로 네 CPU에 제한했고 `LIVE_E2E=0`,
+  `OMK_NO_LOCAL_LLM=1`, `OMK_OFFLINE=1`을 사용했다.
+- 전체 테스트: **8,279 통과, 852 환경·실계정 조건 skip, 실패 0, 종료 0**.
+  WPL 149, agent 870, AI 676, book compiler 22, coding-agent 5,695,
+  protocol 137, TUI 730개가 통과했다. 이 수치는 CI 실행 결과가 아니다.
+- 준비 단계의 TB 106개와 전체 guard 378개 통과를 전체 제품 테스트 수에 다시 더하지 않는다.
+- 최종 후보 확인 중 공유 트리에 Devin 요청 코드·테스트·문서와 그 변경 로그가 별도로
+  바뀐 것을 발견했다. 이 4개 파일의 후속 변경은 제외하고 `003f7b081a`와 릴리스
+  메타데이터만 별도 detached worktree에 옮겼다. 변경 로그의 같은 파일에 섞인 후속
+  항목도 원래 커밋의 본문과 대조해 분리했다. 운영자의 변경·인증은 덮어쓰지 않는다.
+- 격리 후보에서도 전체 테스트 8,279개 통과·852개 조건 skip·실패 0을 재확인했다.
+  공유 트리 실행과 같은 검사를 중복 합산하지 않았다.
+- 격리 후보의 `npm run check`(Node guard 378개 포함), 7개 pack dry-run·공개 entrypoint
+  import, 빌드 CLI의 `--version`·`--help`·`run --help`가 모두 종료 0이었다.
+  모든 pack의 버전은 0.99.0이고, 필수 진입점 누락·비공개 상태 경로는 없었다.
+- 후보 diff의 Gitleaks 검사는 완전 redaction과 기존 규칙으로 종료 0, 탐지 0건이었다.
+- `--release`의 stale-worktree guard는 별도 최종 배포 gate다. 후보를 main에 연결하고
+  이 작업의 임시 체크아웃을 제거한 뒤 기본 checkout에서 실행한다. guard나 이름을
+  바꿔 검사를 피하지 않는다.
+
+## 배포 완료 조건
+
+기존 `build-binaries.yml`의 태그 기반 경로만 사용한다. 로컬 `npm publish`, 인증 변경,
+새로운 실행 권한·모델 호출은 없다. release source와 tag의 동일 SHA 검사를 유지한다.
+후보는 검토한 명시 경로만 stage하고 staged diff 전체를 확인한 뒤 commit·tag·push한다.
+
+공식 workflow가 6개 플랫폼 바이너리 빌드, 검사·테스트, npm 7개 패키지 게시,
+GitHub Release 생성을 완료해야 한다. 최종 판정은 태그의 main 포함,
+GitHub `v0.99.0` Release와 7개 npm `latest`의 일치다. 기존 token 기반 인증을 사용하며
+OIDC/Sigstore provenance는 주장하지 않는다. 실패 시 원인을 확인하고 기존 배포의
+무결성 경계를 유지하며, 완료 전에는 배포 성공이라고 보고하지 않는다.

package/docs/run-protocol.md

@@ -41,7 +41,8 @@ Recovery commands share command IDs and the generation cap; none resets budgets.
 
 `linux-command-dag-v1` adds `RunDagWriter` / `RunDagTask` with 1–16 nodes, explicit
 artifact dependencies, disjoint write scopes and 1–2 preapproved command attempts per
-node. `orderRunDag()` provides deterministic FIFO topological order and
+node. Optional `maxConcurrentTasks` accepts only 1 or 2; omission preserves legacy
+serialization and serial execution. `orderRunDag()` provides deterministic FIFO topological order and
 `runDagAncestors()` includes the full transitive input closure. `RunTaskRetryCommand`
 / `parseRunTaskRetryCommand()` pins a task selection to the original input and exact
 run revision/generation. These pure contracts never authorize dispatch or authenticate
@@ -53,8 +54,8 @@ then supplies authenticated checks to the existing claim-closure reducer.
 This does not replace the task/attempt/evaluation contracts or add execution to
 this package. See [Verified Run](verified-run.md) for the implemented CLI/SDK
 path, candidate recovery and checkpoint-based local writer restart, plus the
-serial command DAG and selective retry, plus remaining live-model, parallel frontier,
-plan amendment and control-surface work.
+bounded command DAG, eager frontier and selective retry, plus remaining live-model,
+verification-edge, plan amendment and control-surface work.
 
 ## Durable goal lifecycle
 

package/docs/runtime-algorithms.md

@@ -154,15 +154,18 @@ Evidence:
 - `packages/coding-agent/test/context-budget-selection-policy-version.test.ts`
 - `packages/coding-agent/test/context-budget-cache-disk.test.ts`
 
-**Working tree:** non-queued native `xai` requests started through `AgentSession.prompt()` now derive a bounded automatic skill grant from live discovered descriptions after ordinary prompt-template expansion. The selector scores task text separately from camelCase-aware path-to-skill-name signals, excludes explicit-only skills, caps automatic matches at three, and adds `headroom` only under lexical or measured context pressure. `AgentSession.prompt()` merges the result with settings/SDK/bang selections only for that request. Queued steering/follow-up messages reuse the active run's system prompt and do not trigger another selection pass.
+**Working tree:** non-queued native `xai` and `devin` requests started through `AgentSession.prompt()` now derive a bounded automatic skill grant from live discovered descriptions after ordinary prompt-template expansion. The selector scores task text separately from camelCase-aware path-to-skill-name signals, excludes explicit-only skills, caps automatic matches at three, and adds `headroom` only under lexical or measured context pressure. `AgentSession.prompt()` merges the result with settings/SDK/bang selections only for that request. Queued steering/follow-up messages reuse the active run's system prompt and do not trigger another selection pass.
 
 Evidence:
 
 - `packages/coding-agent/src/core/active-skill-state.ts`
 - `packages/coding-agent/src/core/skill-selector.ts`
+- `packages/coding-agent/src/core/harness-skills.ts`
 - `packages/coding-agent/src/core/grok-harness.ts`
+- `packages/coding-agent/src/core/devin-harness.ts`
 - `packages/coding-agent/src/core/agent-session.ts`
 - `packages/coding-agent/test/grok-active-skills.test.ts`
+- `packages/coding-agent/test/devin-active-skills.test.ts`
 - `packages/coding-agent/test/skill-selector.property.test.ts`
 
 **Working tree:** context files now treat their global/local relevance baseline