monte3l
OrganizationAWS automation platform: a TypeScript library, the m3l CLI, a 16-script operations fleet, and an operations console — built ESM-only on Node 24 under an AI agent-operating model.
Categories
Indexed Skills (25)
harness-guide
Answers "which skill handles X" and "how do I invoke Y" by reading docs/contributing/skill-routing.md. Reachable only via the explicit /harness-guide command — never auto-invoked from prose, so it costs zero skill-listing budget (see bin/check-context-budget.mjs's collectSkillDescriptions exclusion for this frontmatter key).
auditing
Fan out parallel Explore subagents to audit a topic, aggregate findings against current repo state, ask clarifying questions, enter plan mode with a plan — no code written. Use for /auditing, "audit the codebase", "audit [topic]", "review the current state of", "investigate and plan", or "find gaps in X and plan the fixes".
creating-prs
Verify quality gates, push the branch, open a PR with a Conventional Commit title and body from commit history, then decide and execute its merge path. Use for /creating-prs, "open a PR", "create a pull request", "ship this for review", "get this merged", or after finishing a fix. Requires gh CLI auth. GitHub stance: gh CLI (ADR-0030).
eslint-flat-config
How this repo's ESLint flat config (eslint.config.js) works and how to change it safely — scoping rules, override blocks, `ignores`, plugins, or type-checked lint errors (projectService, TS5110). Use for "eslint is complaining", "turn off this rule for tests", "why is eslint failing", or "add an eslint rule" — anything touching eslint.config.js here.
implementing-scripts
Implement the real business logic of an already-scaffolded consumer script under scripts/<name>/ — the TDD + hub-and-spoke loop turning its contract page into tested steps/ modules. Use for "implement"/"fill in"/"wire up" a script that already has a directory. No directory yet → scaffolding-scripts first; library modules → implementing-submodules instead.
implementing-submodules
Implement a documented-but-empty Core or AWS submodule of @m3l-automation/m3l-common end-to-end from its docs/reference spec, under strict TDD + hub-and-spoke. Use for "implement"/"build"/"flesh out" a submodule that already has a docs/reference/{core,aws}/<name>.md page, even named casually ("the retry stuff"). No spec page yet → scaffolding-submodules first. Docs stance: context7 MCP (ADR-0093).
promoting-work-log-lessons
Closes the work-log → rules loop: reads docs/logs/*.md, finds lessons recurring across logs, promotes them into .claude/rules, agents, or a skill's SKILL.md. Use for /promoting-work-log-lessons, "promote work-log lessons", "which lessons keep recurring", "what does our session telemetry say". m3l MCP stance: adr_query (ADR-0096).
researching-anthropic-guidance
Fan out parallel Explore subagents to search official Anthropic sources only for best practices on a topic, then synthesize consensus and flag contradictions. Use for /researching-anthropic-guidance, "what does Anthropic recommend for X", or mid-task when a decision hinges on Anthropic's position. Not for general library docs or non-Anthropic web research.
resolving-merge-conflicts
Detects an in-progress rebase/merge and resolves real conflicts here: src/**/tests/** logic (handed back), provenance sidecars, package.json dependencies, tracker collisions — gated, pushed signed. Derived-artifact conflicts auto-resolve via ADR-0024 first. Use for /resolving-merge-conflicts, "resolve the merge conflicts", "the rebase is conflicting", "git says CONFLICT", or "help me finish this rebase". Skip for CI failures or bot findings.
resolving-pr-comments
Resolves automated PR review bot findings end-to-end: fetch the bot comment, fix every Must-fix, best-effort Should-fix, fold in Nits already in-region, run gates, reconcile docs, commit, push, reply. Use for "fix what the bot flagged", "claude-pr-review posted FAIL", "address PR feedback". Skip manual reviews/generic CI failures. GitHub stance: hub-only, GitHub MCP (ADR-0030).
reviewing-dependabot-prs
Reviews every open Dependabot PR, proposes merge/hold/reject by semver level and check status, executes the confirmed batch: auto-merge safe bumps, hold risky ones, close bad ones. Use for /reviewing-dependabot-prs, "review the dependabot PRs", "merge the safe dependabot updates". Only acts on explicit invocation after batch confirmation. GitHub stance: gh CLI (ADR-0030). Docs stance: context7 MCP (ADR-0093).
scaffolding-scripts
Scaffold a brand-new script package under scripts/<name>/ via workspace:* — runs pnpm scaffold:script, installs, builds, smoke-runs, hands off to implementing-scripts. Use for "add"/"create"/"scaffold" a script/job/CLI under scripts/ with no directory yet. Library modules → scaffolding-submodules; existing documented submodule → implementing-submodules.
scaffolding-submodules
Scaffold a brand-new Core or AWS submodule with no docs/reference page or src dir yet: module folder, failing placeholder test, barrel re-export, docs stub — hands off to implementing-submodules. Use for "add"/"create"/"scaffold" a new library module, even named casually ("a cache module"). Already has a spec page → implementing-submodules directly.
starting-work
The pre-work decision gate: inspects git state, recommends location, branch (feat/fix <slug>), session name (ADR-0087/0088), PR requirement, push target, and PR sequence (ADR-0072) for multi-unit scope — confirmed before any write. Invoke for "implement", "build", "fix", "refactor", even unnamed. Skip for research/questions.
syncing-docs
Reconciles doc metadata after a submodule/script ships or before release: re-stamps provenance sidecars, regenerates "N of 22" doc-count sites, confirms exports documented, verifies script conformance, regenerates the reference index, runs markdown lint. Use for /syncing-docs, "sync docs", "reconcile docs", "stamp provenance", or after implementing pipelines finish. m3l MCP stance: adr_query (ADR-0096).
triaging-ci
Diagnose a CI failure via gh CLI: resolve the failing run, fetch logs, map the failure to its pipeline step, report root cause plus the exact local repro command, and present 3-5 fix options. Use for /triaging-ci, "why did CI fail", "CI is failing", "debug the CI run", or a specific run ID/URL. GitHub stance: gh CLI (ADR-0030; no Actions tools in the MCP toolset). m3l MCP stance: hooks_query/adr_query (ADR-0096).
triaging-scan-alerts
Triage GitHub code-scanning alerts on a PR/branch via gh CLI: fetch open alerts, group by tool and severity, map to file:line, gate on error-severity CodeQL blocking merge, present remediation — no code edited. Use for /triaging-scan-alerts, "check the CodeQL alerts". Skip for bot findings or CI failures. GitHub stance: gh CLI (ADR-0030).
tsconfig-strict-esm
How this repo's TypeScript project is configured — strict/ESM options in tsconfig.base.json, the build-vs-tooling split — and how to change it safely. Use when editing tsconfig*.json, adding a project reference, or debugging TS5110/TS2834/2835 errors, even phrased as "add a package", "the build can't find my import", or "why do I need .js in imports". Not for generic TypeScript questions.
vitest-coverage-types-mocks
How this repo runs Vitest — coverage gate, v8 per-file thresholds, mocking/type-testing patterns. Use for "coverage is failing", "mock this module", or when editing vitest.config.ts, writing/fixing a mock (vi.mock, vi.spyOn, vi.hoisted), or adding expectTypeOf assertions. Pairs with tests.md. Not for generic Vitest questions unrelated to this repo.
writing-commits
Inspect the staged diff, select the Conventional Commit type, draft an imperative subject ≤70 chars, decide whether a body is needed, structure it as What → Why/How → semver impact, add Co-Authored-By when AI-assisted, then commit. Use for "commit this", "write a commit", "stage and commit".
writing-work-logs
Produces a structured Markdown work log in docs/logs/ covering what shipped, what went as planned, what diverged and why, and durable lessons. Invoked by /writing-work-logs or "document this task", "log what happened", "record what we built". Invoke proactively when a significant task wraps up — real-time context is lost once the session closes.
finishing-work
Runs the post-merge close-out tail creating-prs doesn't: return to main and pull, delete the merged branch, prune stale remote refs, prompt for a tracker flip + sync:hub, prompt for a work log. Use for /finishing-work, "clean up after this PR", "the PR merged, wrap this up", or when a merged branch/worktree/refs linger. GitHub stance: gh CLI (ADR-0030), matching creating-prs.
refreshing-anthropic-guidance
Sweeps the local Claude Code harness — agents, skills, hooks, rules, workflows, CLAUDE.md — against Anthropic's current guidance/models, diffs last-verified state, plans remediation. Use for /refreshing-anthropic-guidance, "is our harness up to date with Anthropic", "model pins current". Distinct from researching-anthropic-guidance and auditing.
typescript-configuration
How this repo's TypeScript project is configured — strict/ESM options in tsconfig.base.json, the build-vs-tooling split — and how to change it safely. Use when editing tsconfig*.json, adding a project reference, or debugging TS5110/TS2834/2835 errors, even phrased as "add a package", "the build can't find my import", or "why do I need .js in imports". Not for generic TypeScript questions.
vitest-testing
How this repo runs Vitest — coverage gate, v8 per-file thresholds, mocking/type-testing patterns. Use for "coverage is failing", "mock this module", or when editing vitest.config.ts, writing/fixing a mock (vi.mock, vi.spyOn, vi.hoisted), or adding expectTypeOf assertions. Pairs with tests.md. Not for generic Vitest questions unrelated to this repo.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.