Parcha-ai
- 14 skills
- 0 followers
- 3 hours ago last updated
- ▌ Tdd · parcha-ai bundleTest-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
- ▌ Recap · parcha-ai bundleReconstruct what a Claude Code or Codex agent actually did during one exact session, including goals, decisions, actions, file and git changes, tests, failures, recoveries, final state, and open work. Use when the user says "/recap", "recap this session", "what did this agent do", "give me the full handoff", or asks for an evidence-backed account of a long-running coding-agent session. Use Recall for prior-session discovery; use Recap after a session is identified. Never use it to expose hidden reasoning or to infer unobserved actions.
- ▌ Autoqa · parcha-ai bundleSelf-QA a repo's running application end-to-end — discover how it runs, build a feature×modality test plan, execute it against a live instance, and produce a witnessed pass/fail report. Use when the user says "autoqa", "QA this repo/branch", "verify every feature works", "release readiness check", "test this like a QA engineer would", or another skill needs an automated QA pass before sign-off.
- ▌ Precap · parcha-ai bundleWrite a precap — the recap of a task as if it were already finished — before or during long-running work, then use it to check whether the run is still on the imagined path. Use when the user says "/precap", "precap this", "write the precap", "what will this take", "imagine this is done", or "are we still on track"; at the start of any task expected to span many steps, hours, or sessions; on resume when a precap.md exists in the working directory; and whenever a long-running agent needs to verify it is doing what it set out to do. It investigates the repo, plan, and git state so the imagined path is grounded rather than guessed. Not a substitute for plan approval or for asking the user about scope.
- ▌ Recall · parcha-ai bundleFind prior Claude Code and Codex sessions with an indexed local search engine, then continue the work, repeat it with fresh inputs, or distill it into a skill. Runs from Claude Code, Codex, or pi; the current index covers Claude Code and Codex transcripts. Use when the user names Recall, says "find that conversation where…", "what did we do last time about…", "continue what we did yesterday on X", "what did codex do on this branch", "turn what we did about Y into a skill", or "remember when you…". Not for searching code (use grep on the repo) or for facts already in MEMORY.md.
- ▌ Tether · parcha-ai bundleTether Slack notifications and replies to the exact Codex, Claude Code, Hermes, or headless session through resumable Hermes conversations. Use when asked to notify Slack, continue coding work from a Slack thread, wire a cron or automation to Slack, or replace direct Slack API calls with session-aware routing.
- ▌ Unslop · parcha-aiCut AI tells from text you write or edit for a human reader (commit messages, PR titles and bodies, docs, code comments, replies). Apply before committing, posting, or sending; leave prose you didn't touch alone.
- ▌ Cascade · parcha-aiCarry large projects through verifiable tasks, parallel work, and durable progress. Use when asked for Cascade, work in loops, or sustained project execution. Skip small edits.
- ▌ Parable · parcha-ai bundleInstall, onboard, and run multi-model coding orchestration or solo single-model mode. Invoke when the user says install Parable, install parable.sh, set up or onboard Parable, connect coding subscriptions, names Parable, hands over a backlog, or asks to route work across models. Solo mode (`parable --solo <model>`) runs with a single model. For first-time install intent, collect subscription choices with the harness's native structured-question tool and then run the bundled parable.sh. Not for a single isolated bugfix or standalone code review.
- ▌ Desloppify · parcha-ai bundleSystematically diagnose and improve whole-codebase maintainability with the official Desloppify engine, combining mechanical scans, blind structural review, root-cause triage, bounded repairs, and regression proof. Use when the user explicitly asks for Desloppify, a codebase health or strict score, a technical-debt cleanup campaign, or a broad maintainability/architecture improvement plan. Do not trigger for an ordinary diff review, a single bug, formatting, or a narrowly scoped refactor.
- ▌ Hands Free · parcha-ai bundlePhone-call bridge for coding agents — when the user is away from the keyboard, the agent calls them (as Unc). Use when the user says "activate hands free" / "deactivate hands free" (also hands-free, handsfree), asks to route questions or approvals to a phone call, or says they're stepping away but want the work to keep moving. Works on Claude Code, Codex, and pi.
- ▌ Review Loop · parcha-ai bundleDrive every reviewer thread on a GitHub pull request to zero unresolved in bounded iterations. Covers bot reviewers such as Greptile and Devin and human reviewers. Fetches every inline comment, review, and issue comment, fixes what is actionable, replies, resolves bot threads through GraphQL, re-requests review once per push, and ends with a fixed-format report. Use when the user says "review loop", "address the review comments", "get this PR to zero unresolved", "greploop", or "make Greptile happy".
- ▌ Blast Radius · parcha-aiFind what a change could break somewhere else before it ships, beyond the diff, and prove the one fact it's safe because of by running real code instead of writing it up. Use for 'blast radius of X', 'what could this break', or reviewing a small diff you don't trust.
- ▌ Verify Release · parcha-ai bundleVerify a specific release in its target environment, follow an in-progress rollout, and resume agreed checks when it is ready. Use for “I deployed it, verify”, “monitor the deployment and continue QA”, or “is this release working in staging/production?”. Discovers the stack and uses live evidence. Not a deployment tool or a full QA campaign by default.