
Retrieval-augmented Q&A over the repo's `docs/` tree. Returns a one-paragraph synthesis plus citations (doc path + heading anchor) and per-citation corroboration verdicts. Field shape and ordering are locked at M3; M5 populates the verdicts.
Audit event-emission posture (structured logging, metrics, trace propagation, event-schema consistency) in a target codebase
Audit lint-posture (linter strictness, formatter presence, CI enforcement, rule-suppression count) in a target codebase
Install consumer-side wirings (Claude Code hooks or .git/hooks/commit-msg) that invoke commventional's trailer-stripping capability. Per ADR-006 §1, the consumer types this slash command — that is the directing act that authorises the write.
Run pronto's composite readiness audit — walks the rubric, delegates per-dimension to installed sibling plugins, falls back to kernel presence checks, emits a composite scorecard
Summarize active plans, open tickets, proposed ADRs, and the latest pulse entry
Full-text search over the repo's `docs/` tree (FTS5-backed). Returns ranked hits with file paths, tags, and matching snippets.
Draft a new plan-scoped ticket from the avanti template into project/tickets/open/
Append a timestamped entry to today's pulse day-file in project/pulse/
Move an artifact forward through its lifecycle and record the transition in pulse
Scaffold a new doc under `docs/` from a Diátaxis template, or update an existing one and bump its `updated:` date. Suggests `## Related` candidates and refreshes the FTS5 index on write.
Show claudit knowledge cache status — freshness, TTL, and domain coverage
Surface (and optionally fix) doc-tree drift — duplicates, dead links, stale docs, template non-compliance, missing `## Related` blocks. Read-only by default; `--apply` does mechanical fixes; `--apply-semantic` emits diffs for human review.
Force-refresh the claudit knowledge cache from official Anthropic documentation
Audit and optimize Claude Code configuration with dynamic best-practice research
Draft a new ADR (Architecture Decision Record) from the avanti template into project/adrs/
Enforces conventional commits, conventional comments, and engineering ownership. Auto-invokes when creating commits, pull requests, or reviewing code.
Show pronto readiness snapshot — last audit score, installed siblings, dimensions below threshold, dimensions not configured
Draft a new plan from the avanti template into project/plans/active/
Audit code-documentation depth (README quality, docs coverage, staleness, internal link health) in a target codebase
Fetch a GitHub PR body, strip engineering-ownership trailers/footers, write the cleaned body back via gh pr edit. Consumer-invoked per-call.
Strip Co-Authored-By trailers and "Generated with/by Claude" footers from a text blob. Read-only capability — emits cleaned text on stdout.
Walk the weakest-scoring rubric dimensions from the last audit and offer per-dimension remediation (install recommended plugin, walk through roll-your-own, or skip)
Scaffold the pronto kernel (AGENTS.md, project/ container, .pronto/, .claude/ seed, .gitignore additions) into the current repo and propose installs for recommended sibling plugins
Run pronto's kernel presence checks — non-delegable filesystem checks (AGENTS.md, project/, .pronto/, README, LICENSE, .gitignore, .claude/)
SDLC hygiene audit — scores plan freshness, ticket hygiene, ADR completeness, pulse cadence; emits pronto wire contract JSON under --json
Bifrost memory system health dashboard
Bifrost memory system setup wizard
Audit and improve an existing Quickstop plugin's quality against Claude Code plugin spec
Audit an existing Claude Code skill's quality with research-first analysis and 6-category scoring
Improve an existing Claude Code skill using audit findings or manual direction
Consolidate inbox transcripts into structured memory — extracts observations and merges into MEMORY.md, journal, and procedures
Retrieve claudit knowledge cache domains (ecosystem, core-config, optimization). Checks freshness and auto-refreshes stale domains.
Search memory — sends Huginn (quick grep) first, then Munin (deep agent search) if needed
Scaffold a new Quickstop plugin with correct structure and conventions
Build a new Claude Code skill from scratch with research-first architecture and opinionated structure