loop-council
Sends an open design question to several independent voices in parallel and returns a decision with its evidence. Use before building anything that touches…
Closes a slice or a session - adversarial audit of the diff, ledger entry, selective commit, honest report, and the handoff for the next session. Use before considering anything finished.
$ npx -y skills add cbdreamer11/CB-loop-kit-claude-plugin --skill loop-close --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/loop-closeContext preview
The summary Claude sees to decide when to auto-load this skill.
Closes a slice or a session - adversarial audit of the diff, ledger entry, selective commit, honest report, and the handoff for the next session. Use before considering anything finished.
name: loop-close description: Closes a slice or a session - adversarial audit of the diff, ledger entry, selective commit, honest report, and the handoff for the next session. Use before considering anything finished.
Closing is not paperwork. It is the step that makes the next session cheap and stops the same work from being done twice.
Dispatch the **`loop-auditor`** agent on the full diff, with the item's non-negotiables from `.loop/STATE.md`. Its job is to refute. Iterate until there are **zero CONFIRMED findings** — and do that *before* anything is published, not after.
If the auditor finds a real bug, also look for **the class of that bug**: search for its siblings elsewhere in the codebase and close them in the same session, or record them as their own items. One fixed instance and three live ones is not a fix.
uncommitted work in the tree.
action, and it stays that way even when it is obviously fine.
means merged, not "it works on my machine".
Anything unverified becomes `GAP: <why>`. Anything blocked becomes `PARKED: <why> + exactly what is missing`.
and how to detect it next time. This file is the compound interest of the method.
Three lists, no softening:
environment). Never dress a gap as done.
permission to publish. State the exact thing needed, not a vague ask.
If something failed, say it failed and show the output. A report that reads better than the work is the one failure this method cannot recover from.
Leave `.loop/STATE.md` so the next session needs no instructions: the next unchecked item is next, and the command to run it is written beside it. Print that command.
A working method for building real software with coding agents, across many sessions.
Sends an open design question to several independent voices in parallel and returns a decision with its evidence. Use before building anything that touches…
Checks that the loop is actually installed and working in this project - files present, role profiles valid, connections alive, effort level real. Use after…
The planning session. Turns a goal into an ordered list of thin complete slices, decides which role and model runs each following session, and writes it all to…
Session 0. Sets up the loop in this project - detects the stack, checks which connections actually work (git, database, browser, payments sandbox), asks the…
Runs this project's verification contract and decides honestly whether something is verified, a gap, or a lie. Use before calling anything done, and whenever a…
The build loop - locate yourself, investigate the real code, build one slice completely, verify it for real, fix and re-verify, then close and hand off. Use…