agent-nuvira 1.74.1 → 1.74.2

package/.agents/skills/code-assessment/SKILL.md

@@ -0,0 +1,44 @@
+---
+name: code-assessment
+description: Perform a structured codebase assessment: read the project, evaluate it across correctness/security/performance/architecture/testability dimensions, produce a gap-findings list, and deliver prioritized recommendations with effort estimates. Use when the goal asks to assess, evaluate, review, audit, or analyze code quality.
+version: 1.0.0
+---
+
+# code-assessment
+
+Perform a structured codebase assessment: read the project, evaluate it across correctness/security/performance/architecture/testability dimensions, produce a gap-findings list, and deliver prioritized recommendations with effort estimates. Use when the goal asks to assess, evaluate, review, audit, or analyze code quality.
+
+## Goal pattern
+
+assess evaluate review audit analyze code quality codebase project architecture security performance correctness gaps recommendations
+
+## Parameters
+
+- target (file-path [default: .]): Path or scope to assess (default: the whole project)
+- focus (string): Optional comma-separated dimension focus (e.g. security, performance)
+
+## Steps
+
+1. [context-gatherer] Map the project first — read the top level (list_dir) and key manifests: package.json / pyproject.toml / go.mod, README, tsconfig / eslint config, and the test layout. Determine:
+- the language/framework stack and entry points
+- the build/test/typecheck commands (from package.json scripts or equivalent)
+- the module boundaries (src/, lib/, app/ dirs) and their sizes
+Record the file count and rough LOC per top-level area so the assessment has scope context.
+
+2. [reviewer] Evaluate the code across these dimensions — read real files (read_file / code_search), never guess:
+- correctness: error handling, edge cases, null/undefined paths, boundary conditions
+- security: secrets in code, injection (SQL/command/HTML), authn/authz gaps, unsafe deserialization, dependency risk
+- performance: obvious O(n^2) patterns, N+1 queries, blocking calls in hot paths, unbounded caches
+- architecture: coupling, god modules, duplicate logic, missing interfaces, config sprawl, dead code
+- testability: coverage gaps, untestable functions (hidden side effects), missing unit/integration boundaries
+For each finding record: file:line, the issue, why it matters, and a severity (critical / major / minor). (after: step-0)
+
+3. [writer] Produce the assessment artifact with this exact structure:
+1. Scope — what was assessed (stack, size, areas)
+2. Gap findings — each with file:line, severity, and why it matters (sorted critical → minor)
+3. Strengths — what is already done well (keep it honest)
+4. Prioritized recommendations — grouped by quick wins / this quarter / later, each with effort (S/M/L) and risk
+5. Suggested next step — the single highest-leverage action to take first
+Deliver it as a structured markdown artifact (heading per section, bullets per finding). (after: step-1)
+
+4. [reviewer] Verify the artifact: every critical/major finding has a file:line reference (no vague "somewhere in the codebase" claims), severities are consistent, and the recommendations are actionable (a concrete change, not a platitude). Revise the artifact if any finding lacks evidence. (after: step-2)

package/.agents/skills/docx/SKILL.md

@@ -0,0 +1,55 @@
+---
+name: docx
+description: Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of Word doc, word document, .docx, .dotx, or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a report, memo, letter, template, or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.
+version: 1.0.0
+---
+
+# docx
+
+Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of Word doc, word document, .docx, .dotx, or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a report, memo, letter, template, or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.
+
+## Goal pattern
+
+docx doc word document template dotx create read edit manipulate convert report memo letter toc tracked changes comments
+
+## Parameters
+
+(none)
+
+## Steps
+
+1. [context-gatherer] Determine the task and choose the approach — a .docx is a ZIP archive of XML files:
+- Create a new document → write a `docx` (npm) script (see the create step for the footguns)
+- Edit an existing document → `unzip` → edit `word/document.xml` → re-zip (docx-js cannot open existing files)
+- Read content → `pandoc -t markdown file.docx`
+Gather the deliverable spec: document type (report / memo / letter / template), target format (.docx vs .dotx), US Letter vs A4 page size, and whether tracked changes or comments are required.
+
+2. [runner] Create a new document with docx-js. `docx` is preinstalled — do NOT run `npm install` first; write the script and `require('docx')` directly. Only if that require fails: `npm install docx`. Known footguns:
+- Page size defaults to A4. For US Letter set `page: { size: { width: 12240, height: 15840 } }` (DXA; 1440 = 1″)
+- Landscape: pass portrait dimensions and `orientation: PageOrientation.LANDSCAPE` — docx-js swaps width/height internally
+- Tables need dual widths: set `columnWidths` on the table AND `width` on every cell, both in `WidthType.DXA` (PERCENTAGE breaks in Google Docs); column widths must sum to the table width
+- Table shading: use `ShadingType.CLEAR`, never `SOLID` (renders black)
+- Lists: never insert `•` literally; use a `numbering` config with `LevelFormat.BULLET`
+- `ImageRun` requires `type:` ("png", "jpg", …)
+- `PageBreak` must be inside a `Paragraph`
+- Never use `
+` — use separate `Paragraph` elements
+- TOC: headings must use built-in `HeadingLevel.*`; custom heading styles need `outlineLevel` set or they won't appear
+- Don't use a table as a horizontal rule — use a paragraph bottom border instead
+- Dot-leader / right-aligned-on-same-line: use `PositionalTab` (`alignment: PositionalTabAlignment.RIGHT`, `leader: PositionalTabLeader.DOT`) inside a `TextRun`, not literal `.` or space padding (after: step-0)
+
+3. [runner] Edit an existing document. Legacy `.doc` files must be converted first: `python scripts/office/soffice.py --headless --convert-to docx file.doc`. Then:
+1. `unzip -q doc.docx -d unpacked/`
+2. `find unpacked -type l -delete` — strip symlink entries; docx from external parties is untrusted
+3. `python scripts/merge_runs.py unpacked/` — coalesce fragmented runs so text is findable. Word splits text across many `<w:r>` runs (revision ids, spell-check markers), so a phrase you can see often doesn't exist as a contiguous string in the XML; merge_runs merges adjacent identically-formatted runs without changing content or rendering (it also accepts a `.docx` directly: `python scripts/merge_runs.py doc.docx -o merged.docx`)
+4. Edit `unpacked/word/document.xml` in place — do NOT reformat or pretty-print
+5. Re-zip: `(cd unpacked && rm -f ../out.docx && zip -Xr ../out.docx .)`
+6. Validate: `python scripts/office/validate.py out.docx --original doc.docx` — XSD checks; `--auto-repair` fixes common issues (after: step-0)
+
+4. [runner] Tracked changes and comments. Redlining: validate with `--author "<the name you redlined under>"` (needs `--original`) — it reports any text you changed without a `<w:ins>`/`<w:del>` around it, which is easy to do by accident and invisible in the accepted view. Wrap runs in `<w:ins>`/`<w:del>` with `w:id`, `w:author`, `w:date` attributes. Inside `<w:del>`, the text element is `<w:delText>`, not `<w:t>`. A deleted paragraph mark (`<w:pPr><w:rPr><w:del w:id=".." w:author=".." w:date=".."/></w:rPr></w:pPr>`) means merge this paragraph into the next — deleting a paragraph outright is that plus a `<w:del>` around every run. The `<w:del/>` must come before the rPr's other children; their order is schema-enforced. To produce a clean copy with all tracked changes accepted: `python scripts/accept_changes.py in.docx out.docx`. Caveat: accepting a deleted paragraph mark should join that paragraph to the one below it, but `accept_changes.py` and `pandoc --track-changes=accept` don't always — they strip the deleted text but leave the emptied paragraph behind (a stray empty bullet when it was auto-numbered); `pandoc` never joins the paragraphs, `accept_changes.py` joins them correctly except when the deleted paragraph is followed by an empty spacer paragraph. An empty bullet in either view is an artifact of that view, not a defect in the document — check paragraph deletions in the XML.
+
+Comments require six cross-linked files — use the helper. Directory mode when you'll also be editing `document.xml` (saves an unzip/rezip cycle): `python scripts/comment.py unpacked/ "Fees & expenses cap is too low"`, `python scripts/comment.py unpacked/ "Agreed" --parent 0`. Against a `.docx` directly: `python scripts/comment.py contract.docx "This cap is too low" -o annotated.docx`. The script writes `comments.xml`, `commentsExtended.xml`, `commentsIds.xml`, `commentsExtensible.xml`, the relationships, and the content-type overrides; comment IDs are auto-assigned. It then prints the `<w:commentRangeStart>`/`<w:commentRangeEnd>`/`<w:commentReference>` snippet to add to `word/document.xml` so the comment anchors to specific text — until you place those markers, the comment exists but is not visible. (after: step-2)
+
+5. [reviewer] Verify the output — render it and look at it:
+`python scripts/office/soffice.py --headless --convert-to pdf output.docx` then `pdftoppm -jpeg -r 100 output.pdf page` then `ls page-*.jpg` and Read the images (`pdftoppm` zero-pads page numbers to the width of the page count: `page-01.jpg`…`page-12.jpg`).
+Run `python scripts/office/validate.py out.docx` (XSD checks) after any edit. Dependencies: `docx` (npm, preinstalled — install only if `require('docx')` fails), `pandoc`, LibreOffice (`soffice`), `pdftoppm` (Poppler). (after: step-1, step-2, step-3)

package/.agents/skills/index.json

@@ -0,0 +1,105 @@
+{
+  "version": 1,
+  "updatedAt": "2026-08-18T08:39:06.004Z",
+  "skills": [
+    {
+      "name": "website-deploy",
+      "description": "Deploy a static site or built web app to a hosting provider (Cloudflare Pages, Netlify, Vercel, GitHub Pages, AWS S3+CloudFront, Azure Static Web Apps, Firebase Hosting) and verify the live URL. Use when the goal asks to deploy, publish, host, or ship a website or web app to a hosting provider.",
+      "version": "1.0.0",
+      "author": "agent-nuvira",
+      "tags": [
+        "deploy",
+        "website",
+        "hosting",
+        "cloudflare",
+        "netlify",
+        "vercel",
+        "github-pages",
+        "aws",
+        "azure",
+        "firebase",
+        "publish"
+      ],
+      "source": "agent-nuvira/bundled",
+      "updatedAt": "2026-08-18T08:39:06.004Z"
+    },
+    {
+      "name": "code-assessment",
+      "description": "Perform a structured codebase assessment: read the project, evaluate it across correctness/security/performance/architecture/testability dimensions, produce a gap-findings list, and deliver prioritized recommendations with effort estimates. Use when the goal asks to assess, evaluate, review, audit, or analyze code quality.",
+      "version": "1.0.0",
+      "author": "agent-nuvira",
+      "tags": [
+        "assessment",
+        "code-review",
+        "audit",
+        "quality",
+        "gaps",
+        "recommendations"
+      ],
+      "source": "agent-nuvira/bundled",
+      "updatedAt": "2026-08-18T08:39:06.004Z"
+    },
+    {
+      "name": "technical-roadmap",
+      "description": "Build a phased technical roadmap from the current state to a target state: capture the current architecture, define the target, then produce ordered phases with dependencies, effort, risk, and success criteria. Use when the goal asks for a roadmap, migration plan, technical plan, or phased upgrade path.",
+      "version": "1.0.0",
+      "author": "agent-nuvira",
+      "tags": [
+        "roadmap",
+        "planning",
+        "migration",
+        "architecture",
+        "phases",
+        "strategy"
+      ],
+      "source": "agent-nuvira/bundled",
+      "updatedAt": "2026-08-18T08:39:06.004Z"
+    },
+    {
+      "name": "plan-create-track",
+      "description": "Plan and track a multi-step job: break the goal into ordered, verifiable steps, declare them with plan_todo, work through them updating status (running → done, or blocked with a note), and finish with a summary. Use for any job with 2+ steps where progress visibility matters.",
+      "version": "1.0.0",
+      "author": "agent-nuvira",
+      "tags": [
+        "plan",
+        "todo",
+        "tracking",
+        "checklist",
+        "progress",
+        "execution"
+      ],
+      "source": "agent-nuvira/bundled",
+      "updatedAt": "2026-08-18T08:39:06.004Z"
+    },
+    {
+      "name": "test-strategy",
+      "description": "Plan and run a deep test pass: map the test surface, choose the right matrix (unit / integration / e2e, focused runs for changed code), execute the real commands (npm test, vitest, pytest, etc.), and deliver a verdict with evidence. Use when the goal asks to test, verify, check for regressions, or prove a change is safe.",
+      "version": "1.0.0",
+      "author": "agent-nuvira",
+      "tags": [
+        "test",
+        "testing",
+        "verification",
+        "regression",
+        "coverage",
+        "quality"
+      ],
+      "source": "agent-nuvira/bundled",
+      "updatedAt": "2026-08-18T08:39:06.004Z"
+    },
+    {
+      "name": "docx",
+      "description": "Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of Word doc, word document, .docx, .dotx, or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a report, memo, letter, template, or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.",
+      "version": "1.0.0",
+      "author": "agent-nuvira",
+      "tags": [
+        "docx",
+        "word",
+        "documents",
+        "office"
+      ],
+      "source": "agent-nuvira/bundled",
+      "updatedAt": "2026-08-18T08:39:06.004Z"
+    }
+  ]
+}

package/.agents/skills/plan-create-track/SKILL.md

@@ -0,0 +1,33 @@
+---
+name: plan-create-track
+description: Plan and track a multi-step job: break the goal into ordered, verifiable steps, declare them with plan_todo, work through them updating status (running → done, or blocked with a note), and finish with a summary. Use for any job with 2+ steps where progress visibility matters.
+version: 1.0.0
+---
+
+# plan-create-track
+
+Plan and track a multi-step job: break the goal into ordered, verifiable steps, declare them with plan_todo, work through them updating status (running → done, or blocked with a note), and finish with a summary. Use for any job with 2+ steps where progress visibility matters.
+
+## Goal pattern
+
+plan create track steps todo checklist progress multi-step job execute work through order
+
+## Parameters
+
+- goal (string (required)): The goal to plan and execute
+
+## Steps
+
+1. [planner] Break the goal into 3–7 ordered steps. Each step must be:
+- independently verifiable (you will know it is done by running/reading something)
+- small enough to complete in one working session
+- ordered so dependencies come first (but no stricter than necessary)
+Give each step a short stable id (reproduce, fix, verify) and a one-line description. Declare them with the plan_todo tool (action: create, goal + steps).
+
+2. [runner] Work the steps IN ORDER, updating plan_todo (action: update, id + status) at each transition:
+- mark the current step running before starting it
+- mark it done only after its exit criterion is verified (a test passed, a file reads correctly, a command succeeded)
+- if a step is blocked (an external dependency, a failure you cannot fix now), mark it blocked and CONTINUE with the next step — never stall the whole plan on one step
+Use the coding tools (read_file / edit_file / run_terminal) to actually do the work — the plan is a map, not the work itself. (after: step-0)
+
+3. [reviewer] Close the loop: confirm every step is done or explicitly blocked, re-check the exit criteria for the done steps (run the final verification once more if cheap), and write a closing summary: what was completed, what is blocked (if anything), and the natural next step. (after: step-1)

package/.agents/skills/technical-roadmap/SKILL.md

@@ -0,0 +1,33 @@
+---
+name: technical-roadmap
+description: Build a phased technical roadmap from the current state to a target state: capture the current architecture, define the target, then produce ordered phases with dependencies, effort, risk, and success criteria. Use when the goal asks for a roadmap, migration plan, technical plan, or phased upgrade path.
+version: 1.0.0
+---
+
+# technical-roadmap
+
+Build a phased technical roadmap from the current state to a target state: capture the current architecture, define the target, then produce ordered phases with dependencies, effort, risk, and success criteria. Use when the goal asks for a roadmap, migration plan, technical plan, or phased upgrade path.
+
+## Goal pattern
+
+roadmap migration plan technical plan phased upgrade path target state current state phases dependencies milestones
+
+## Parameters
+
+- target (file-path [default: .]): Path or scope for the roadmap (default: the whole project)
+- horizon (string [default: 6 months]): Roadmap horizon, e.g. "6 months" or "this quarter" (default: 6 months)
+
+## Steps
+
+1. [context-gatherer] Establish the CURRENT state with evidence: read the architecture-relevant files (entry points, config, module boundaries, package manifests), and note the stack, key flows, and known constraints (legacy pieces, hard dependencies, team-visible risk). Keep it factual — cite files.
+
+2. [writer] Define the TARGET state as concrete outcomes, not slogans: for each area (stack, architecture, quality, operations) state what changes and what the measurable success criterion is (e.g. "typecheck passes with noEmit", "deploys are under 5 minutes"). (after: step-0)
+
+3. [planner] Design the PHASES between current and target. Each phase must have:
+- a clear outcome and exit criteria
+- dependencies on other phases (explicit dependsOn)
+- effort estimate (S/M/L) and risk (low/medium/high)
+- what is explicitly OUT of scope (so phases stay small and shippable)
+Order them so each phase leaves the system working (never a long broken window). Typically 3–5 phases. (after: step-1)
+
+4. [writer] Produce the roadmap artifact: current state (cited) → target state (measurable) → phases (each with outcome, dependencies, effort, risk, out-of-scope), plus a critical-path note (which phases gate everything else) and a first-step recommendation. (after: step-2)

package/.agents/skills/test-strategy/SKILL.md

@@ -0,0 +1,42 @@
+---
+name: test-strategy
+description: Plan and run a deep test pass: map the test surface, choose the right matrix (unit / integration / e2e, focused runs for changed code), execute the real commands (npm test, vitest, pytest, etc.), and deliver a verdict with evidence. Use when the goal asks to test, verify, check for regressions, or prove a change is safe.
+version: 1.0.0
+---
+
+# test-strategy
+
+Plan and run a deep test pass: map the test surface, choose the right matrix (unit / integration / e2e, focused runs for changed code), execute the real commands (npm test, vitest, pytest, etc.), and deliver a verdict with evidence. Use when the goal asks to test, verify, check for regressions, or prove a change is safe.
+
+## Goal pattern
+
+test verify regression check coverage suite unit integration e2e pass run tests prove safe
+
+## Parameters
+
+- scope (choice [default: full]): Scope of the test pass: focused | unit | integration | full (default: full)
+- target (file-path): Path or files to focus on (for scope=focused)
+
+## Steps
+
+1. [context-gatherer] Map the test surface with evidence: read the test config (vitest/jest/pytest config in the manifests), list the test files (glob "**/*.test.*" or tests/), and note the commands that run them (from package.json scripts or equivalent). Identify:
+- the unit test entry points and how fast they are
+- any integration/e2e suites and their prerequisites (services, fixtures, env vars)
+- which areas changed and deserve a focused run first
+
+2. [planner] Choose the matrix — never just "run everything":
+- focused: the tests touching the changed code (fastest, run FIRST)
+- unit: the full unit suite (the main regression net)
+- integration: the suites that exercise real boundaries (DB, HTTP, filesystem)
+- e2e: the slow end-to-end flows — run only when unit+integration are green
+- typecheck/build: static verification alongside the tests
+Record the exact commands and the order, with the reason each level matters. (after: step-0)
+
+3. [runner] Execute the matrix IN ORDER with the real commands via run_terminal (never claim tests pass without running them):
+- start with the focused run — a failure here tells you the change broke something before the slow suites waste time
+- then the full unit suite
+- then integration (and e2e only if the cheaper levels are green)
+- run typecheck/build as the static gate
+On a failure: read the failing test + the code (read_file), fix, and re-run that focused test until green before moving up the matrix. (after: step-1)
+
+4. [reviewer] Deliver the verdict with evidence: what ran (exact commands + pass/fail counts), what passed, what failed and why (file:line), coverage gaps that matter, and a clear recommendation (safe to merge / needs fixes / needs more tests). Never state "tests pass" without the actual run output. (after: step-2)

package/.agents/skills/website-deploy/SKILL.md

@@ -0,0 +1,46 @@
+---
+name: website-deploy
+description: Deploy a static site or built web app to a hosting provider (Cloudflare Pages, Netlify, Vercel, GitHub Pages, AWS S3+CloudFront, Azure Static Web Apps, Firebase Hosting) and verify the live URL. Use when the goal asks to deploy, publish, host, or ship a website or web app to a hosting provider.
+version: 1.0.0
+---
+
+# website-deploy
+
+Deploy a static site or built web app to a hosting provider (Cloudflare Pages, Netlify, Vercel, GitHub Pages, AWS S3+CloudFront, Azure Static Web Apps, Firebase Hosting) and verify the live URL. Use when the goal asks to deploy, publish, host, or ship a website or web app to a hosting provider.
+
+## Goal pattern
+
+deploy publish host website web app site landing page cloudflare pages netlify vercel github pages hosting static
+
+## Parameters
+
+- provider (choice (required)): Hosting provider to deploy to: cloudflare-pages, netlify, vercel, github-pages, aws-s3, azure-swa, or firebase-hosting
+- projectName (string): Provider project/site/bucket name
+- outputDir (file-path [default: .]): Directory containing the built site files (default: current directory)
+- productionBranch (string [default: main]): Production branch for the hosting project (default: main)
+
+## Steps
+
+1. [context-gatherer] Inspect the site directory to identify the built output (index.html or a framework build folder like dist/ or build/), detect which hosting CLIs are installed and authenticated on this machine (wrangler, netlify, vercel, gh, aws, firebase), and report the site type and output directory.
+
+2. [runner] Ensure the hosting project is ready for the target provider (skip creation if the project already exists):
+- cloudflare-pages: wrangler 4 does NOT auto-create Pages projects, so create it first (ignore the error if it already exists): Run `wrangler pages project create {{projectName}} --production-branch {{productionBranch}}`
+- netlify: no setup needed — the CLI creates the site on first deploy
+- vercel: no setup needed — the CLI creates the project on first deploy
+- github-pages: ensure the folder is a git repo with a remote and a .nojekyll file: Run `git init 2>/dev/null; touch .nojekyll; git add -A && git commit -m "init" 2>/dev/null || true`
+- aws-s3: ensure the S3 bucket exists with static website hosting enabled: Run `aws s3api create-bucket --bucket {{projectName}} --region us-east-1 2>/dev/null || true; aws s3 website s3://{{projectName}} --index-document index.html --error-document 404.html 2>/dev/null || true`
+- azure-swa: no setup needed — the CLI creates the app
+- firebase-hosting: ensure firebase.json exists with a public directory set to {{outputDir}} (create it if missing). (after: step-0)
+
+3. [runner] Deploy the built site in '{{outputDir}}' to the '{{provider}}' provider and capture the deployment URL printed by the command:
+- cloudflare-pages: Run `wrangler pages deploy {{outputDir}} --project-name {{projectName}}`
+- netlify: Run `npx netlify-cli deploy --prod --dir={{outputDir}}`
+- vercel: Run `npx vercel --prod --yes`
+- github-pages: Run `git add -A && git commit -m "deploy: website" && git push origin {{productionBranch}}` (or `npx gh-pages -d {{outputDir}}` to push a gh-pages branch)
+- aws-s3: Run `aws s3 sync {{outputDir}} s3://{{projectName}} --delete` (deployment URL is https://{{projectName}}.s3-website-<region>.amazonaws.com)
+- azure-swa: Run `npx @azure/static-web-apps-cli deploy --env production`
+- firebase-hosting: Run `npx firebase-tools deploy --only hosting` (after: step-1)
+
+4. [runner] Verify the deployed site is live: fetch the deployment URL printed by the deploy step with curl and confirm it returns HTTP 200 and the site HTML. For example: Run `curl -s -o /dev/null -w "%{http_code}" <deployed-url>` then Run `curl -s <deployed-url> | head -c 400` (after: step-2)
+
+5. [reviewer] Review the deployment: confirm the live URL returns HTTP 200 with the expected page content, and that the deployed files (HTML/CSS/JS) are present and correctly referenced. (after: step-3)

package/README.md

@@ -47,7 +47,7 @@ This table highlights core capabilities for quick machine parsing and comparison
 | Security guardrails | Privacy-focused; PII detection; prompt-injection defenses; security scan CLI |
 | Publishing | Standalone eject & npm publishing (`npx agent-nuvira`) |
 | **Routing strategy** | **Thompson-sampling bandit + uncertainty escalation + per-model learning + promotion gate A/B + routing rules + hard constraints** |
-| **Test suite** | **3,806 tests across 149 files — 100% passing** |
+| **Test suite** | **4,864 tests across 217 files — 100% passing** |
 | **Vector backend** | **Native FAISS (automatic), pure-JS IVF fallback, exact JSON fallback** |
 
 
@@ -58,9 +58,9 @@ This table highlights core capabilities for quick machine parsing and comparison
   layer: a deterministic rule fast-path (<5ms, works offline) with LLM verification behind it,
   collapsing to a single action every command consumes identically. Before any work begins,
   Agent-Nuvira prints a 🧠 Understanding card — the resolved goal, target, scope, acceptance
-  criteria, and risk flags — so the request is **seen before it is run**. Neither Freebuff nor
-  Hermes does this. The live activity board then renders each agent step as it happens —
-  no black box, no silently half-understood goals
+  criteria, and risk flags — so the request is **seen before it is run**. The live activity
+  board then renders each agent step as it happens — no black box, no silently
+  half-understood goals
 - **🛠️ Execution that repairs itself** — a failed task is not just reported: failures are
   classified, automatically repaired with escalating retries, independently verified after the
   fix, and the lesson is stored in persistent failure memory so the same mistake is less likely
@@ -105,7 +105,7 @@ This table highlights core capabilities for quick machine parsing and comparison
 - **VS Code extension** — Chat Panel with streaming responses, slash commands, and session history;
   Diagnostic → AI Fix from lightbulb menu; Code Lens actions (Test/Review/Explain/Fix) above functions
   and classes; 9 commands, inline code suggestions, diff viewer, agent progress panel
-- **22-platform multi-channel gateway** — `buff gateway start / send / status / alias` runs the agent from Telegram, Discord, Slack, WhatsApp (Cloud API + Baileys personal bridge), Email, Signal, DingTalk, Feishu, WeCom, Mattermost, Matrix, generic Webhook, BlueBubbles (iMessage), ntfy, Teams, Google Chat, Weixin, SMS (Twilio), IRC (two-way), SimpleX (two-way), and Home Assistant — all opt-in via the same env vars Hermes uses, with a guaranteed delivery ledger (auto-retry on `buff gateway start`), the dashboard Channels send-test, and a `X/22 platforms configured` status line
+- **22-platform multi-channel gateway** — `buff gateway start / send / status / alias` runs the agent from Telegram, Discord, Slack, WhatsApp (Cloud API + Baileys personal bridge), Email, Signal, DingTalk, Feishu, WeCom, Mattermost, Matrix, generic Webhook, BlueBubbles (iMessage), ntfy, Teams, Google Chat, Weixin, SMS (Twilio), IRC (two-way), SimpleX (two-way),  and Home Assistant — all opt-in via standard env vars per platform, with a guaranteed delivery ledger (auto-retry on `buff gateway start`), the dashboard Channels send-test, and a `X/22 platforms configured` status line
 - **Remote agent federation** — multi-machine collaboration with protocol, server, and client
 - **Web UI dashboard** — React dashboard with DAG visualization, model health, cost charts, and history browser
 - **Hybrid model routing** — intelligent model selection based on task complexity, cost, and availability
@@ -2040,7 +2040,8 @@ npx tsc --noEmit
 
 | Version | Date | Key Changes |
 |---------|------|-------------|
-| **v1.63.0** | Aug 2026 | **Hermes messaging campaign — 22-platform multi-channel gateway** — the full Hermes `gateway/` surface with env-var parity: DingTalk, Feishu, WeCom, Mattermost, Matrix, generic Webhook, BlueBubbles, ntfy, Teams, Google Chat, Weixin (iLink bot API), SMS (Twilio), IRC (RFC 1459 net/tls, two-way), SimpleX (daemon WS, two-way), Home Assistant (REST) on top of Telegram/Discord/Slack/WhatsApp — same `TWILIO_*`/`IRC_*`/`SIMPLEX_*`/`HASS_*` creds as Hermes, zero new SDK deps; WhatsApp Baileys personal bridge (`buff whatsapp pair`); guaranteed delivery ledger with auto-retry; dashboard Channels send-test; `buff gateway status` platform-count line; SimpleX inbound (auto-accept + allowlists + reconnect); IRC inbound (addressing + allowlist + PING/PONG + reconnect); 4,280 tests |
+| **v1.74.2** | Aug 2026 | **Dashboard-first — chat is the front door (Phases 1–8 delivered)** — token-streaming typewriter, spec-complete markdown + syntax highlight, artifact cards (diff per-file accept/reject + commit-accepted, result, deploy — copy buttons + keyboard nav), inline command-run execution cards, project attach with code-map context, persisted sessions with smart-rail resume/search/rename/delete, composer attachments (file/paste/drag), /learn skill authoring + marketplace install, streaming cancel + retry; 4,864 root + 263 dashboard tests; CLI engine unchanged |
+| **v1.63.0** | Aug 2026 | **Messaging campaign — 22-platform multi-channel gateway** — the full gateway surface with env-var-compatible credentials: DingTalk, Feishu, WeCom, Mattermost, Matrix, generic Webhook, BlueBubbles, ntfy, Teams, Google Chat, Weixin (iLink bot API), SMS (Twilio), IRC (RFC 1459 net/tls, two-way), SimpleX (daemon WS, two-way), Home Assistant (REST) on top of Telegram/Discord/Slack/WhatsApp — service-native `TWILIO_*`/`IRC_*`/`SIMPLEX_*`/`HASS_*` creds, zero new SDK deps; WhatsApp Baileys personal bridge (`buff whatsapp pair`); guaranteed delivery ledger with auto-retry; dashboard Channels send-test; `buff gateway status` platform-count line; SimpleX inbound (auto-accept + allowlists + reconnect); IRC inbound (addressing + allowlist + PING/PONG + reconnect); 4,280 tests |
 | **v1.0.0** | Apr 2026 | Initial release — Core CLI with chat, 5 built-in providers (expandable to 17+ via plugins), config, models |
 | **v1.1.0** | Apr 2026 | Model discovery with search/filter |
 | **v1.2.0** | Apr 2026 | AI-assisted file editing (edit command) |

package/dist/agents/agent.d.ts

@@ -10,8 +10,8 @@ import type { InferenceOptions } from '../config/types.js';
 export type TaskStatus = 'pending' | 'running' | 'completed' | 'failed';
 /**
  * H2 — one sub-agent delegation spec on a delegate task step. The agent runs
- * this sub-agent with a FRESH isolated context (Freebuff `spawn_agents` /
- * Hermes `delegate_tool.py` parity).
+ * this sub-agent with a FRESH isolated context (`spawn_agents` /
+ * `delegate_tool.py` parity).
  */
 export interface TaskDelegation {
     /** The specialized sub-agent type (registered in the ModuleRegistry). */

package/dist/agents/module-registry.js

@@ -275,7 +275,7 @@ export class ModuleRegistry {
             isBuiltin: true,
         });
         // H2 — delegate agent: fans out taskStep.delegation sub-agent specs in
-        // parallel (Freebuff spawn_agents / Hermes delegate_tool.py parity).
+        // parallel (spawn_agents / delegate_tool.py parity).
         registry.register('delegate', () => new DelegateAgent(), {
             name: 'Delegate',
             description: 'Fans out sub-agent delegation specs in parallel with fresh isolated contexts and aggregates their summary results',

package/dist/agents/module-registry.js.map

@@ -1 +1 @@
-{"version":3,"file":"module-registry.js","sourceRoot":"","sources":["../../src/agents/module-registry.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAGH,OAAO,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AACnD,OAAO,EAAE,oBAAoB,EAAE,MAAM,8BAA8B,CAAC;AACpE,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AACrD,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AACrD,OAAO,EAAE,QAAQ,EAAE,MAAM,uBAAuB,CAAC;AACjD,OAAO,EAAE,WAAW,EAAE,MAAM,0BAA0B,CAAC;AACvD,OAAO,EAAE,YAAY,EAAE,MAAM,2BAA2B,CAAC;AACzD,OAAO,EAAE,kBAAkB,EAAE,MAAM,kCAAkC,CAAC;AACtE,OAAO,EAAE,aAAa,EAAE,MAAM,4BAA4B,CAAC;AAC3D,OAAO,EAAE,gBAAgB,EAAE,MAAM,0BAA0B,CAAC;AAC5D,OAAO,EAAE,QAAQ,EAAE,MAAM,uBAAuB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,6BAA6B,CAAC;AAC5D,OAAO,EAAE,gBAAgB,EAAE,MAAM,gCAAgC,CAAC;AAClE,OAAO,EAAE,qBAAqB,EAAE,MAAM,qCAAqC,CAAC;AAC5E,OAAO,EAAE,aAAa,EAAE,MAAM,4BAA4B,CAAC;AAC3D,OAAO,EAAE,WAAW,EAAE,UAAU,EAAE,MAAM,+BAA+B,CAAC;AAsBxE,8CAA8C;AAC9C,MAAM,OAAO,mBAAoB,SAAQ,KAAK;IAC5C,YAAY,SAAiB;QAC3B,KAAK,CAAC,yCAAyC,SAAS,GAAG,CAAC,CAAC;QAC7D,IAAI,CAAC,IAAI,GAAG,qBAAqB,CAAC;IACpC,CAAC;CACF;AAED,+EAA+E;AAE/E;;;;;;;;;;;GAWG;AACH,MAAM,OAAO,cAAc;IACzB,kDAAkD;IAC1C,SAAS,GAAG,IAAI,GAAG,EAAwB,CAAC;IACpD,0CAA0C;IAClC,QAAQ,GAAG,IAAI,GAAG,EAA0B,CAAC;IACrD,sDAAsD;IAC9C,QAAQ,CAAW;IAE3B,YAAY,QAAmB;QAC7B,IAAI,CAAC,QAAQ,GAAG,QAAQ,IAAI,WAAW,EAAE,CAAC;IAC5C,CAAC;IAED,wEAAwE;IAExE;;;;;;;;OAQG;IACH,QAAQ,CACN,SAAiB,EACjB,OAAqB,EACrB,IAA+E;QAE/E,IAAI,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC;YAClC,MAAM,IAAI,KAAK,CACb,eAAe,SAAS,6DAA6D,CACtF,CAAC;QACJ,CAAC;QACD,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;QACvC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,EAAE;YAC3B,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,WAAW,EAAE,IAAI,CAAC,WAAW;YAC7B,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,0BAA0B,EAAE;YACxD,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,EAAE,iBAAiB,CAAC,CAAC;IACxB,CAAC;IAED;;;OAGG;IACH,kBAAkB,CAChB,SAAiB,EACjB,OAAqB,EACrB,IAA+E;QAE/E,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;QACvC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,EAAE;YAC3B,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,WAAW,EAAE,IAAI,CAAC,WAAW;YAC7B,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,0BAA0B,EAAE;YACxD,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,EAAE,iBAAiB,CAAC,CAAC;IACxB,CAAC;IAED;;;OAGG;IACH,UAAU,CAAC,SAAiB;QAC1B,MAAM,UAAU,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC;QACpD,IAAI,CAAC,QAAQ,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC;QAEhC,IAAI,UAAU,EAAE,CAAC;YACf,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,4BAA4B,EAAE;gBAC1D,SAAS;aACV,EAAE,iBAAiB,CAAC,CAAC;QACxB,CAAC;QAED,OAAO,UAAU,CAAC;IACpB,CAAC;IAED,wEAAwE;IAExE;;;;;;OAMG;IACH,SAAS,CAAC,SAAiB;QACzB,MAAM,OAAO,GAAG,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;QAC9C,IAAI,CAAC,OAAO,EAAE,CAAC;YACb,MAAM,IAAI,mBAAmB,CAAC,SAAS,CAAC,CAAC;QAC3C,CAAC;QACD,OAAO,OAAO,EAAE,CAAC;IACnB,CAAC;IAED;;OAEG;IACH,SAAS,CAAC,SAAiB;QACzB,OAAO,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;IACvC,CAAC;IAED;;;OAGG;IACH,WAAW,CAAC,SAAiB;QAC3B,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;IACtC,CAAC;IAED;;OAEG;IACH,WAAW,CAAC,MAA0C;QACpD,MAAM,GAAG,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,MAAM,EAAE,CAAC,CAAC;QAC/C,OAAO,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC;IAC3C,CAAC;IAED;;OAEG;IACH,OAAO,CAAC,SAAiB;QACvB,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,IAAI,IAAI,IAAI,CAAC;IACpD,CAAC;IAED;;OAEG;IACH,IAAI,IAAI;QACN,OAAO,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC;IAC7B,CAAC;IAED,wEAAwE;IAExE;;OAEG;IACH,MAAM,CAAC,kBAAkB,CAAC,QAAmB;QAC3C,MAAM,QAAQ,GAAG,IAAI,cAAc,CAAC,QAAQ,CAAC,CAAC;QAE9C,+DAA+D;QAC/D,QAAQ,CAAC,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE,CAAC,IAAI,YAAY,EAAE,EAAE;YACrD,IAAI,EAAE,SAAS;YACf,WAAW,EAAE,0DAA0D;YACvE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE,CAAC,IAAI,oBAAoB,EAAE,EAAE;YACtE,IAAI,EAAE,kBAAkB;YACxB,WAAW,EAAE,kDAAkD;YAC/D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,sDAAsD;YACnE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,+DAA+D;YAC5E,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,6CAA6C;YAC1D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,uCAAuC;YACpD,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,uDAAuD;YACpE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,KAAK,EAAE,GAAG,EAAE,CAAC,IAAI,QAAQ,EAAE,EAAE;YAC7C,IAAI,EAAE,KAAK;YACX,WAAW,EAAE,6CAA6C;YAC1D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,6DAA6D;YAC1E,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE,CAAC,IAAI,YAAY,EAAE,EAAE;YACrD,IAAI,EAAE,SAAS;YACf,WAAW,EAAE,iDAAiD;YAC9D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,gBAAgB,EAAE,GAAG,EAAE,CAAC,IAAI,kBAAkB,EAAE,EAAE;YAClE,IAAI,EAAE,gBAAgB;YACtB,WAAW,EAAE,wDAAwD;YACrE,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,8DAA8D;YAC3E,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE,CAAC,IAAI,gBAAgB,EAAE,EAAE;YAC9D,IAAI,EAAE,aAAa;YACnB,WAAW,EAAE,qDAAqD;YAClE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,KAAK,EAAE,GAAG,EAAE,CAAC,IAAI,QAAQ,EAAE,EAAE;YAC7C,IAAI,EAAE,KAAK;YACX,WAAW,EAAE,mEAAmE;YAChF,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,WAAW,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACxD,IAAI,EAAE,WAAW;YACjB,WAAW,EAAE,gEAAgE;YAC7E,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE,CAAC,IAAI,gBAAgB,EAAE,EAAE;YAC9D,IAAI,EAAE,cAAc;YACpB,WAAW,EAAE,wEAAwE;YACrF,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,mBAAmB,EAAE,GAAG,EAAE,CAAC,IAAI,qBAAqB,EAAE,EAAE;YACxE,IAAI,EAAE,mBAAmB;YACzB,WAAW,EAAE,qGAAqG;YAClH,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,uEAAuE;QACvE,qEAAqE;QACrE,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,mHAAmH;YAChI,IAAI,EAAE,MAAM;YACZ,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,OAAO,QAAQ,CAAC;IAClB,CAAC;CACF;AAED,+EAA+E;AAE/E,IAAI,eAAe,GAA0B,IAAI,CAAC;AAElD;;;;;;GAMG;AACH,MAAM,UAAU,iBAAiB;IAC/B,IAAI,CAAC,eAAe,EAAE,CAAC;QACrB,eAAe,GAAG,cAAc,CAAC,kBAAkB,EAAE,CAAC;IACxD,CAAC;IACD,OAAO,eAAe,CAAC;AACzB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,mBAAmB;IACjC,eAAe,GAAG,IAAI,CAAC;AACzB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,iBAAiB,CAAC,QAAwB;IACxD,MAAM,QAAQ,GAAG,eAAe,CAAC;IACjC,eAAe,GAAG,QAAQ,CAAC;IAC3B,OAAO,QAAQ,CAAC;AAClB,CAAC"}
+{"version":3,"file":"module-registry.js","sourceRoot":"","sources":["../../src/agents/module-registry.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAGH,OAAO,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AACnD,OAAO,EAAE,oBAAoB,EAAE,MAAM,8BAA8B,CAAC;AACpE,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AACrD,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AACrD,OAAO,EAAE,QAAQ,EAAE,MAAM,uBAAuB,CAAC;AACjD,OAAO,EAAE,WAAW,EAAE,MAAM,0BAA0B,CAAC;AACvD,OAAO,EAAE,YAAY,EAAE,MAAM,2BAA2B,CAAC;AACzD,OAAO,EAAE,kBAAkB,EAAE,MAAM,kCAAkC,CAAC;AACtE,OAAO,EAAE,aAAa,EAAE,MAAM,4BAA4B,CAAC;AAC3D,OAAO,EAAE,gBAAgB,EAAE,MAAM,0BAA0B,CAAC;AAC5D,OAAO,EAAE,QAAQ,EAAE,MAAM,uBAAuB,CAAC;AACjD,OAAO,EAAE,aAAa,EAAE,MAAM,6BAA6B,CAAC;AAC5D,OAAO,EAAE,gBAAgB,EAAE,MAAM,gCAAgC,CAAC;AAClE,OAAO,EAAE,qBAAqB,EAAE,MAAM,qCAAqC,CAAC;AAC5E,OAAO,EAAE,aAAa,EAAE,MAAM,4BAA4B,CAAC;AAC3D,OAAO,EAAE,WAAW,EAAE,UAAU,EAAE,MAAM,+BAA+B,CAAC;AAsBxE,8CAA8C;AAC9C,MAAM,OAAO,mBAAoB,SAAQ,KAAK;IAC5C,YAAY,SAAiB;QAC3B,KAAK,CAAC,yCAAyC,SAAS,GAAG,CAAC,CAAC;QAC7D,IAAI,CAAC,IAAI,GAAG,qBAAqB,CAAC;IACpC,CAAC;CACF;AAED,+EAA+E;AAE/E;;;;;;;;;;;GAWG;AACH,MAAM,OAAO,cAAc;IACzB,kDAAkD;IAC1C,SAAS,GAAG,IAAI,GAAG,EAAwB,CAAC;IACpD,0CAA0C;IAClC,QAAQ,GAAG,IAAI,GAAG,EAA0B,CAAC;IACrD,sDAAsD;IAC9C,QAAQ,CAAW;IAE3B,YAAY,QAAmB;QAC7B,IAAI,CAAC,QAAQ,GAAG,QAAQ,IAAI,WAAW,EAAE,CAAC;IAC5C,CAAC;IAED,wEAAwE;IAExE;;;;;;;;OAQG;IACH,QAAQ,CACN,SAAiB,EACjB,OAAqB,EACrB,IAA+E;QAE/E,IAAI,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC;YAClC,MAAM,IAAI,KAAK,CACb,eAAe,SAAS,6DAA6D,CACtF,CAAC;QACJ,CAAC;QACD,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;QACvC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,EAAE;YAC3B,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,WAAW,EAAE,IAAI,CAAC,WAAW;YAC7B,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,0BAA0B,EAAE;YACxD,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,EAAE,iBAAiB,CAAC,CAAC;IACxB,CAAC;IAED;;;OAGG;IACH,kBAAkB,CAChB,SAAiB,EACjB,OAAqB,EACrB,IAA+E;QAE/E,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;QACvC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,EAAE;YAC3B,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,WAAW,EAAE,IAAI,CAAC,WAAW;YAC7B,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,0BAA0B,EAAE;YACxD,SAAS;YACT,IAAI,EAAE,IAAI,CAAC,IAAI;YACf,SAAS,EAAE,IAAI,CAAC,SAAS,IAAI,KAAK;SACnC,EAAE,iBAAiB,CAAC,CAAC;IACxB,CAAC;IAED;;;OAGG;IACH,UAAU,CAAC,SAAiB;QAC1B,MAAM,UAAU,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC;QACpD,IAAI,CAAC,QAAQ,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC;QAEhC,IAAI,UAAU,EAAE,CAAC;YACf,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,4BAA4B,EAAE;gBAC1D,SAAS;aACV,EAAE,iBAAiB,CAAC,CAAC;QACxB,CAAC;QAED,OAAO,UAAU,CAAC;IACpB,CAAC;IAED,wEAAwE;IAExE;;;;;;OAMG;IACH,SAAS,CAAC,SAAiB;QACzB,MAAM,OAAO,GAAG,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;QAC9C,IAAI,CAAC,OAAO,EAAE,CAAC;YACb,MAAM,IAAI,mBAAmB,CAAC,SAAS,CAAC,CAAC;QAC3C,CAAC;QACD,OAAO,OAAO,EAAE,CAAC;IACnB,CAAC;IAED;;OAEG;IACH,SAAS,CAAC,SAAiB;QACzB,OAAO,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;IACvC,CAAC;IAED;;;OAGG;IACH,WAAW,CAAC,SAAiB;QAC3B,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;IACtC,CAAC;IAED;;OAEG;IACH,WAAW,CAAC,MAA0C;QACpD,MAAM,GAAG,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,MAAM,EAAE,CAAC,CAAC;QAC/C,OAAO,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC;IAC3C,CAAC;IAED;;OAEG;IACH,OAAO,CAAC,SAAiB;QACvB,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,IAAI,IAAI,IAAI,CAAC;IACpD,CAAC;IAED;;OAEG;IACH,IAAI,IAAI;QACN,OAAO,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC;IAC7B,CAAC;IAED,wEAAwE;IAExE;;OAEG;IACH,MAAM,CAAC,kBAAkB,CAAC,QAAmB;QAC3C,MAAM,QAAQ,GAAG,IAAI,cAAc,CAAC,QAAQ,CAAC,CAAC;QAE9C,+DAA+D;QAC/D,QAAQ,CAAC,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE,CAAC,IAAI,YAAY,EAAE,EAAE;YACrD,IAAI,EAAE,SAAS;YACf,WAAW,EAAE,0DAA0D;YACvE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE,CAAC,IAAI,oBAAoB,EAAE,EAAE;YACtE,IAAI,EAAE,kBAAkB;YACxB,WAAW,EAAE,kDAAkD;YAC/D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,sDAAsD;YACnE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,+DAA+D;YAC5E,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,6CAA6C;YAC1D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,uCAAuC;YACpD,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,uDAAuD;YACpE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,KAAK,EAAE,GAAG,EAAE,CAAC,IAAI,QAAQ,EAAE,EAAE;YAC7C,IAAI,EAAE,KAAK;YACX,WAAW,EAAE,6CAA6C;YAC1D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE,CAAC,IAAI,WAAW,EAAE,EAAE;YACnD,IAAI,EAAE,QAAQ;YACd,WAAW,EAAE,6DAA6D;YAC1E,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE,CAAC,IAAI,YAAY,EAAE,EAAE;YACrD,IAAI,EAAE,SAAS;YACf,WAAW,EAAE,iDAAiD;YAC9D,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,gBAAgB,EAAE,GAAG,EAAE,CAAC,IAAI,kBAAkB,EAAE,EAAE;YAClE,IAAI,EAAE,gBAAgB;YACtB,WAAW,EAAE,wDAAwD;YACrE,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,8DAA8D;YAC3E,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE,CAAC,IAAI,gBAAgB,EAAE,EAAE;YAC9D,IAAI,EAAE,aAAa;YACnB,WAAW,EAAE,qDAAqD;YAClE,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,KAAK,EAAE,GAAG,EAAE,CAAC,IAAI,QAAQ,EAAE,EAAE;YAC7C,IAAI,EAAE,KAAK;YACX,WAAW,EAAE,mEAAmE;YAChF,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,WAAW,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACxD,IAAI,EAAE,WAAW;YACjB,WAAW,EAAE,gEAAgE;YAC7E,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE,CAAC,IAAI,gBAAgB,EAAE,EAAE;YAC9D,IAAI,EAAE,cAAc;YACpB,WAAW,EAAE,wEAAwE;YACrF,IAAI,EAAE,KAAK;YACX,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,QAAQ,CAAC,QAAQ,CAAC,mBAAmB,EAAE,GAAG,EAAE,CAAC,IAAI,qBAAqB,EAAE,EAAE;YACxE,IAAI,EAAE,mBAAmB;YACzB,WAAW,EAAE,qGAAqG;YAClH,IAAI,EAAE,IAAI;YACV,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,uEAAuE;QACvE,qDAAqD;QACrD,QAAQ,CAAC,QAAQ,CAAC,UAAU,EAAE,GAAG,EAAE,CAAC,IAAI,aAAa,EAAE,EAAE;YACvD,IAAI,EAAE,UAAU;YAChB,WAAW,EAAE,mHAAmH;YAChI,IAAI,EAAE,MAAM;YACZ,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QAEH,OAAO,QAAQ,CAAC;IAClB,CAAC;CACF;AAED,+EAA+E;AAE/E,IAAI,eAAe,GAA0B,IAAI,CAAC;AAElD;;;;;;GAMG;AACH,MAAM,UAAU,iBAAiB;IAC/B,IAAI,CAAC,eAAe,EAAE,CAAC;QACrB,eAAe,GAAG,cAAc,CAAC,kBAAkB,EAAE,CAAC;IACxD,CAAC;IACD,OAAO,eAAe,CAAC;AACzB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,mBAAmB;IACjC,eAAe,GAAG,IAAI,CAAC;AACzB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,iBAAiB,CAAC,QAAwB;IACxD,MAAM,QAAQ,GAAG,eAAe,CAAC;IACjC,eAAe,GAAG,QAAQ,CAAC;IAC3B,OAAO,QAAQ,CAAC;AAClB,CAAC"}

package/dist/agents/orchestrator.js

@@ -843,7 +843,7 @@ export class Orchestrator {
             }
             // Runner and sandbox agents need exclusive access (no parallel).
             // Conservative parallelism (recommended): independent tasks — gatherers,
-            // writers, reviewers — run in PARALLEL within a batch (Freebuff-style),
+            // writers, reviewers — run in PARALLEL within a batch,
             // while tester/debugger/runner (and any strategy-marked serial step) run
             // one at a time because they share files, commands, ports, and sandboxes.
             const exclusiveAgentTypes = ['tester', 'debugger', 'runner'];