ClosedLoop.ai
Resources

Commands reference

Every slash command exposed by the ClosedLoop.ai Claude Code plugins.

Commands are namespaced by plugin (/<plugin>:<command>).

/bootstrap:start

Also /bootstrap:agent-bootstrap. Generate a project-specific suite of expert agents from the current repo.

Flags:

  • --target-command <name>
  • --depth quick|medium|deep
  • --focus <area>
  • --output-dir <path>
  • --dry-run
  • --interactive
  • --update
  • --minimal / --enhance
  • --add-domain <domain>
  • --strategy backup|skip|overwrite|interactive

See Bootstrap a repo.

/code:code

Begin a ClosedLoop coding session against a PRD or plan.

Form:

claude /code:code [working-directory] [--prompt <name>] [--prd <file>] [--plan <file>]

See PRD to plan to execution.

/code:amend-plan

Interactively amend a running plan. Classifies your input as directive, question, confirmation, or unstructured.

Form:

claude /code:amend-plan --workdir <path> --message "<text>" [--state-file <path>]

/code:cancel-code

Stop a running loop by removing its state file.

/code:plan-with-codex

Run a Claude-vs-Codex debate loop to refine a plan.

Form:

claude /code:plan-with-codex [--max-rounds N] [--plan-file PATH] [--codex-model MODEL] "<prompt>"

Defaults: --max-rounds 15, --codex-model gpt-5.4.

See Refine a plan with Codex.

/code-review:start

Multi-agent partitioned code review.

Form:

claude /code-review:start [scope] [--github] [--hygiene-only] [--base <ref>] [--since-last-review] [--full-review]

Scopes:

  • no arg — review current branch vs main
  • staged — review staged changes
  • file1 file2 ... — review specific files
  • 123 — review PR #123

See Code review.

/self-learning:process-learnings

Classify pending learnings, validate, dedupe, and merge into org-patterns.toon. Prunes low performers.

claude /self-learning:process-learnings [workdir]

/self-learning:export-closedloop-learnings

Merge ClosedLoop-specific pending learnings into ~/.closedloop-ai/learnings/closedloop-learnings.json.

claude /self-learning:export-closedloop-learnings [workdir]

/self-learning:push-learnings

Export local patterns to the shared team repo. Requires CLAUDE_ORG_ID.

See Push and pull learnings.

/self-learning:pull-learnings

Import organization patterns. Echo prevention skips patterns originating from the current project.

/self-learning:prune-learnings

Manual pruning per retention.yaml.

/self-learning:goal-stats

Pass rate, top patterns, and trends. Requires at least 5 runs.

Skills (no slash command; invoked by agents or by @judges:run-judges)

  • @judges:run-judges --artifact-type plan|code|prd — run judge batches.

Top-level .claude/commands/

  • /update-documentation — generates CHANGELOG.md and README.md updates. Must run before pushing plugin changes (enforced by .githooks/pre-push).

Plugin marketplace

claude /plugin marketplace install closedloop
claude plugin list

On this page