claudecodex

loom-code

Optional using-loom-code discovery routing and five stations carry one change from plan to PR with non-blocking cross-vendor suggestions, content-bound verification, one closing review, and a GitHub-enforced PR floor. On-demand engineering research resolves decision-critical evidence gaps. Claude Code, Codex + Antigravity CLI.

0 on GitHubby koukov3.72.0MITupdated today
/

Plugin loom-code from marketplace loom in kouko/loom-plugins. Commands read from the repo's manifest — today.

This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.

What it installs

Skills (7)

  • closing-review

    Run closing review and generate an attestation. Use after Build completes or functional changes invalidate prior review evidence.

  • engineering-research

    Resolve evidence gaps in Loom tasks or standalone engineering discussions when an unresolved premise could change a decision, including a small capability or version check. Routine lookups without a consequential gap remain ordinary work.

  • external-review

    Execute one consented independent review through Codex, Claude Code, or Antigravity CLI with an explicit model and effort. The calling review skill keeps its own criteria and verdict checks.

  • maintain

    Route alerts, CI or regression failures, or dogfood incidents into an intent. Use for incidents outside an active unmerged change that arrive without a user's request to fix a bug; that request starts at the entry station.

  • ship

    Publish a reviewed branch and verify PR checks. Use after closing-review, normally with a matching attestation.

  • using-loom-code

    Route general Loom implementation or engineering research requests to the right skill. Use when asked to use Loom without naming a skill.

  • write-plan

    Turn a confirmed intent into implementation tasks. Use when an intent needs a plan; also the entry for a new feature or bug-fix request when loom-design is not installed.

Subagents (4)

  • acceptance-tester

    Plugin-level acceptance-tester agent for loom-code. Dispatched fresh-context by the closing-review station to build and run the change in a clean environment and walk every Acceptance line of the intent, producing docs/loom/<change-id>/acceptance-test-report.md — the document the user reads to accep

  • adversary

    Plugin-level adversary agent for loom-code. Dispatched fresh-context by the build station after all tasks land to make the change fail — mutation or fuzz tooling when the repo declares it, else executable abuse and boundary cases; red-team for a spec, temptation and near-miss attempts for a skill or

  • implementer

    Plugin-level implementer agent for loom-code. Dispatched by the build station for one task of a plan under the engineering baseline — failing test first, one commit for the task. Produces code + tests + a status report (DONE / DONE_WITH_CONCERNS / NEEDS_CONTEXT / BLOCKED), never a verdict. Reusable

  • reviewer

    Fresh-context reviewer for code, docs, spec, spec+adversarial, design, principles, skill, and plan lenses at required spec, pre-build plan, or branch-end checkpoints. Produces PASS / PASS_WITH_NOTES / NEEDS_REVISION with scores and anchored findings; never edits what it reviews. Reusable via subagen

Hooks (6)

  • Stop

    "${CLAUDE_PLUGIN_ROOT}/scripts/loom-python" --ready delivery --hook --host claude; loom_hook_exit_code=$?; if [ "$loom_hook_exit_code" -ne 0 ]; then echo 'loom: delivery unconfirmed; rerun the origina

  • SessionStart

    "${CLAUDE_PLUGIN_ROOT}/hooks/session-start"

  • SessionStart

    if python3 -c 'import json,sys; assert sys.version_info >= (3,7)' >/dev/null 2>&1; then exec python3 "${CLAUDE_PLUGIN_ROOT}/hooks/language-anchor.py"; fi; "${CLAUDE_PLUGIN_ROOT}/scripts/loom-python" -

  • PreToolUse

    if [ -f "${CLAUDE_PLUGIN_ROOT}/scripts/loom_checker.py" ]; then "${CLAUDE_PLUGIN_ROOT}/scripts/loom-python" --ready push --hook; loom_hook_exit_code=$?; if [ "$loom_hook_exit_code" -eq 86 ]; then echo

  • PostToolUse

    if python3 -c 'import json,sys; assert sys.version_info >= (3,7)' >/dev/null 2>&1; then exec python3 "${CLAUDE_PLUGIN_ROOT}/hooks/language-anchor.py"; fi; "${CLAUDE_PLUGIN_ROOT}/scripts/loom-python" -

  • UserPromptSubmit

    if python3 -c 'import json,sys; assert sys.version_info >= (3,7)' >/dev/null 2>&1; then exec python3 "${CLAUDE_PLUGIN_ROOT}/hooks/language-anchor.py"; fi; "${CLAUDE_PLUGIN_ROOT}/scripts/loom-python" -

View source on GitHub

More from kouko/loom-plugins

All 3 plugins →

loom-workflow

claudecodex

Optional using-loom-workflow discovery routing and Loom workflow tools around the loom stations — persistent Outcome Maps, git memory, loom-memory (recall / record / reconcile / retire durable repository lessons), critique (proposal | complexity), recap, handoff, project-retro and skill-retro, CoT explainers, cross-executor second opinion, plus the standalone goal-create and dbt-model-style skills.

14 skills · hooks

0

loom-design

claudecodex

Optional using-loom-design discovery routing and two stations — capture-intent, write-spec — turn a rough idea into a confirmed intent and a risk-declared spec, route non-blocking second-vendor suggestions, and provide product-principles, design-system, architecture-design and design-research tools. Claude Code, Codex + Antigravity CLI.

7 skills

0

Similar plugins

Collection of example skills demonstrating various capabilities including skill creation, MCP building, visual design, algorithmic art, internal communications, web testing, artifact building, Slack GIFs, and theme styling

19 skills

anthropics/skills 193,780

agent-skills

claudecodex

Production-grade engineering skills covering every phase of software development: spec, plan, build, verify, review, and ship.

25 skills · 4 agents

addyosmani/agent-skills 116,782

storybook

claudecodex

Build, preview, and test UI components with Storybook.

7 skills

storybookjs/storybook 91,552

Statically analyze a local JS/Electron app you're authorized to study, with file+line evidence, and build an original reimplementation of what you find. Experimental. Phase 1: JS/Electron static analysis. Phase 2: native ELF/PE/Mach-O container inspection. Phase 3: .NET assembly metadata. Phase 4: EVM bytecode inventory. Phase 5: CUDA fatbin, cubin and PTX inventory. Parses, never executes. Dynamic analysis is a future phase, not yet implemented.

1 skill · 1 command · 1 agent

ruvnet/ruflo 81,589

pake

claudecodex

Package any website or local web build into a lightweight desktop app with Pake CLI.

1 skill

tw93/Pake 63,463

Build powerful, stateful CLI interfaces for any GUI application using the cli-anything harness methodology.

5 commands

HKUDS/CLI-Anything 58,665

Maintain loom-code?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins3 plugins