break-glass
01Skill Claude CodeCodex
Emergency IAM elevation for oncall/debugging. Discovers GCP projects and guides through granting/revoking permissions.
Curated library of Claude Code subagents and skills (23 agents, 20 skills) with an idempotent installer that symlinks into ~/.claude. Frontmatter-driven; spec-compliant with Anthropic and Gemini schemas.
Skill Claude CodeCodex
Emergency IAM elevation for oncall/debugging. Discovers GCP projects and guides through granting/revoking permissions.
Skill Claude CodeCodex
Generate a bottom-up diagram of actual code coupling from cross-references — what the code does, not what was intended. Use for structural PR review; pair with diagram-gen-spec for intended-vs-actual drift.
Skill Claude CodeCodex
Generate a top-down diagram of intended architecture from a mache schema file — what the PR proposes, not what the code does. Local reviewer context only; pair with diagram-gen-emergent for spec-vs-reality drift.
Skill Claude CodeCodex
Find internal docs (not for external consumers), review against code reality, bead real items, archive stale files. Converts markdown debt into tracked beads.
Skill Claude CodeCodex
Review a PR, ticket, or path and explain it so it sticks — mache/CODEOWNERS/ ticket context synthesized into a taught explanation, not a bare findings list. Own work gets onboarding framing; a peer's gets findings-teaching. Not for posting reviews (pr-review) or the reviewer-only pre-flight alone (review-prep). Local…
Skill Claude CodeCodex
Research-first feature implementation workflow. Use when building new features or net-new code that should follow existing codebase patterns. Enforces pattern discovery before coding; dispatches to repo-specific standards skills for convention checks.
Skill Claude CodeCodex
Before writing glue, a wrapper, or a "small helper", audit whether the capability already exists first-party — in the binary, the codebase, or the platform — and whether existing glue bypasses a better first-party implementation. Use when adding shell/Taskfile/CI glue, when a helper feels trivially writable, or when…
Skill Claude CodeCodex
Compact a session into a handoff at the work SEAM — where this session ends and the next resumes. Writes a rosary mailbox entry, bead comment, jj checkpoint, and /tmp view; --resume drains it next session. Use at the end of a task another session must continue, or to resume one — not a general summary.
Skill Claude CodeCodex
Use for Linear tickets assigned to you — oncall queue sweeps, aged P1/P2 escalations in Triage > 7 days, stale Todos needing cleanup, or an external escalation (Slack/support) surfacing a ticket needing a triaged reply.
Skill Claude CodeCodex
Start or locate a running mache MCP server for a source path or .db file. Builds a .db from source if needed. Base skill used by diagram-gen-spec, diagram-gen-emergent, and review-prep. Returns port and session ID.
Skill Claude CodeCodex
Answers "how are my PRs?", "did I address the review?", "anything waiting on me?". Lectio-first via authored-PR activity (reviews/comments by others, excluding your own replies), falls back to gh for coverage gaps, joins to local worktrees, prints a board of what needs you. Read-only.
Skill Claude CodeCodex
Self-contained playbook for rigorous PR review: state inspection, design-intent hunting, verification rules (cite sources, walk commits, verify artifacts), agent dispatch, falsifiable-matrix output, GitHub posting only with explicit authorization. Generic, sequential phases.
Skill Claude CodeCodex
Give or respond to PR reviews with structural awareness via mache. Review mode: review-prep diagrams + impact analysis, parallel analysis agents, one consolidated GitHub review. Respond mode: address reviewer comments.
Skill Claude CodeCodex
Produce cited, structured comparisons between a project and external systems solving adjacent problems; updates a prior-art matrix. Use for prior-art research or "are we reinventing X?" before committing to a design. Max 5 systems/invocation; no uncited claims.
Skill Claude CodeCodex
Decompose an aspiration into dispatchable bead specs via 5-whys descent + a 7-property rubric. Use when starting new platform/capability/migration work that needs breaking down before agents can ship it, or to audit a backlog for orphan-dispatchable beads.
Skill Claude CodeCodex
Restructure a README for human skimmers and LLM ingestion — fixes frontloaded internals, moves the value prop into the first 200 tokens, generates a diff for review. Preserves content, reorders disclosure. Sub-skill of repo-seo-curator.
Skill Claude CodeCodex
Audit and polish the GitHub repo description — the pitch shown in search results, previews, and embeds. Proposes a stronger one from README + signals; applies via gh repo edit after approval. Sub-skill of repo-seo-curator.
Skill Claude CodeCodex
Audit and set the GitHub repo homepage URL — the link on the repo card's "About" panel and every search result. Detects candidates from docs sites, registries, and demos; applies via gh repo edit after approval. Sub-skill of repo-seo-curator.
Skill Claude CodeCodex
Audit the GitHub OpenGraph social preview image shown when a repo URL is posted on HN/Twitter/Slack/Discord. Flags if missing, checks clarity at small sizes, suggests sources, reports manual upload steps. Sub-skill of repo-seo-curator.
Skill Claude CodeCodex
Audit and apply GitHub topic tags for discoverability — proposes tags from code signals, deduplicates against current tags, applies via gh api after explicit approval. Sub-skill of repo-seo-curator.
Skill Claude CodeCodex
Synthesize a PR's spec-driven diagram, emergent diagram, and impact analysis into reviewer context: what's proposed vs. what the code does vs. blast radius of changed symbols. Local only, never posted to GitHub.
Skill Claude CodeCodex
Fan out isolated, one-per-PR reviews across your queue (own context + worktree per PR, no cross-pollution), then synthesize one report. Pulls the queue from a review inbox (zen) or gh; posts nothing without per-PR authorization.
Skill Claude CodeCodex
Pre-review self-audit of a PR branch: dead struct fields, rotting comments, duplicate types, scope drift — what a human catches but a compiler won't — then dispatches an adversarial agent for deeper design flaws. Use before tagging a reviewer, or after a self-directed refactor.
Skill Claude CodeCodex
Deterministic PR-review chain with pluggable structural intelligence. Fetches the PR into an isolated worktree, runs blast-radius/impact analysis through whatever code-intelligence tool is available (a depgraph tool like modmap, or an MCP server like mache), applies the pr-review-kit discipline, teaches the result via…