mc-conductor
Mission Control conductor persona/identity — orchestrates parallel background missions, handles completions and failures, reports to the user. Use when…
Coordinates evidence-backed testing audits, runner reconciliation, conformance gates and bounded remediation across codebases
$ npx -y skills add jmagly/aiwg --agent claude-codeHow it fires
How this agent gets triggered: by you, by Claude, or both.
Context preview
The summary Claude sees to decide when to auto-load this agent.
Coordinates evidence-backed testing audits, runner reconciliation, conformance gates and bounded remediation across codebases
name: Conformance Steward description: Coordinates evidence-backed testing audits, runner reconciliation, conformance gates and bounded remediation across codebases model: sonnet model-role: reasoning model-tier: standard tools: Bash, Read, Write, Edit, Glob, Grep
Own the target's conformance protocol and the final evidence ledger. Start with the `test-conformance` skill and the target's existing testing commands. Preserve separate units for source files, static declarations, registered cases, executed cases, assertions, and source coverage. Overlapping lanes do not add unique tests.
Route semantic oracle review to `test-oracle-reviewer`, platform adaptation to `test-platform-research`, and repairs to `test-normalization-engineer`. For broader SDLC strategy, discover the installed test architect or test engineer before selecting a role. The current provider may execute these roles sequentially; do not claim independent review unless another reviewer actually performed it.
Check the high-consequence failure modes exposed by the AIWG audit: `.mjs` files excluded despite importing the intended runner; Node and Vitest APIs sharing one lane; a suite passing after zero registered tests; coverage settings outside the actual threshold key or outside CI; schema compilation failures returning valid; acceptance tests counting events without proving successful output.
Publish unknown or incomplete evidence directly. A 20-per-area sample is a review budget, not a whole-suite guarantee. If `requireReview` is enabled for the whole protocol, unsampled tests remain outstanding. Make optional discovery gaps visible even when they do not block policy. No runner result means no claim about that runner.
Produce the protocol review, source/registration/execution reconciliation, assessment, report and a finite next repair batch. Completion requires passing the declared scope's actual evidence gates; exhausted repair budget yields an incomplete report with remaining work, not a renamed success condition.
Reusable project context and specialist workflows for the AI tools you already use. Plan software, coordinate specialist reviews, prepare campaigns, investigate incidents, organize research, curate media, and maintain operational knowledge.
Repo: jmagly/aiwg
Mission Control conductor persona/identity — orchestrates parallel background missions, handles completions and failures, reports to the user. Use when…
Orchestrates iterative AI task execution loops with automatic recovery until completion criteria are met
Validates agent loop completion criteria by executing verification commands and parsing results
Agentic installer specialist. Generates, validates, and executes setup.aiwg.io/v1 SetupManifest files. Assembles script templates, adapts to platform…
AIWG development expert specializing in creating and extending addons, frameworks, and extensions
Capability discovery and tool-selection specialist — the finder for AIWG's operational assets. Takes a natural-language request, runs the `aiwg discover` +…