claude

observe

Measures real harness usage (Claude Code + Codex hooks, Antigravity backfill, opt-in git hooks) into a per-repo .tracer/ store and turns it into improvement proposals with an explicit feedback loop

1 on GitHubby Dokkabei97v2.1.2Apache-2.0updated 16 days agoharness
/

Plugin observe from marketplace dokkabei-harness in Dokkabei97/dokkabei-harness. Commands read from the repo's manifest — 13 days ago.

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

Slash commands (2)

  • /observe:observe-git-hooks

    observe 커밋 경계 수집용 git 훅 opt-in 설치·제거·상태 확인 — post-commit(commit 이벤트 + detached 분석기 → .tracer/commits/<sha>.md, LLM 은 OBSERVE_LLM=1 일 때만·일일 상한), 비차단 pre-commit(미처리 추천 결정론 1줄 stderr, 항상 기존 훅 종료코드), 선택 prepare-commit-msg trailer(--trailer). 기존 훅은 체이닝·백업, core.hooksPath/husky/lefthook 감지 시 안내만. Installs

  • /observe:observe-report

    계측 기반 하네스 건강 리포트 — .tracer 트레이스(Claude Code 훅, Codex·Antigravity 백필)를 결정론 집계(스킬·에이전트 호출, user/model trigger, 로드 수, 턴 결과 committed·cancelled·corrected 후보·failed, 미사용 자산, stale/archive 라이프사이클 후보, 미발화 후보 턴, 교정 후보 쌍, BM25 추천 missed·tune·orchestrate)한 뒤 해석해 description 튜닝 초안·폐기 후보·계측 개선 제안을 만들고, 추천 ID 마다

Hooks (9)

  • SessionStart

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-session.js"

  • UserPromptSubmit

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-prompt.js"

  • PreToolUse

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-skill.js"

  • PreToolUse

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-agent.js"

  • PostToolUse

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-result.js"

  • PostToolUse

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-tool.js"

  • PostToolUseFailure

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-tool.js"

  • Stop

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-stop.js"

  • SessionEnd

    node "${CLAUDE_PLUGIN_ROOT}/bin/hooks/adapters/claude/trace-session.js"

View source on GitHub dokkabei97.github.io

More from Dokkabei97/dokkabei-harness

All 14 plugins →

workflow

claude

Development lifecycle workflows: project rules (AGENTS.md stack facts + specialist core for Claude Code, Codex and Antigravity; Claude path-scoped rules), spec/planning/shipping guides, MR review, handoff, release notes, retrospectives, and CLAUDE.md sync

11 skills · 3 commands · hooks

1

Shared backend specialist layer: PostgreSQL, Kafka, Valkey and distributed-call pitfalls, migration safety, DTO/serialization rules, plus user-invoked API design, docs, test, and perf commands

4 skills · 8 commands · 2 agents

1

search

claude

Elasticsearch specialist layer: ES/Lucene performance, relevance (Korean nori), and vector/hybrid search pitfall hubs, review commands, and a user-invoked agent-team orchestrator

4 skills · 5 commands · 4 agents

1

analyze

claude

Comprehensive static analysis of code quality, performance, architecture, and SQL

4 skills · 4 commands · 3 agents

1

wiki-ops

claude

Operates a cited llm-wiki knowledge vault through gated feed, audit, curation-loop, and Q&A

7 skills · 3 agents

1

harness

claude

Meta-plugin for harness design (v2 execution modes), evolution, generation, validation, and lightweight loops

4 skills · 4 commands · hooks

1

Similar plugins

ecc

claudecodex

Harness-native ECC operator layer - 71 agents, 314 skills, 98 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses

200 skills · 98 commands · 71 agents · 1 MCP server · hooks

affaan-m/ECC 302,088

Comprehensive toolkit for developing Claude Code plugins. Includes 7 expert skills covering hooks, MCP integration, commands, agents, and best practices. AI-assisted plugin creation and validation.

7 skills · 1 command · 3 agents

anthropics/claude-code 155,476

CMEM Pro

claude

CMEM Pro for the Claude app - hooks stream tool use to cmem.ai and inject observations into new sessions and agents

3 skills · 1 MCP server · hooks

thedotmack/claude-mem 116,618

Agentic QE (external: maintained in proffesor-for-testing/agentic-qe, pinned to v3.15.17): a quality-engineering fleet with 11 QE agents, 9 skills, 9 commands, the agentic-qe MCP server pinned to the plugin version, and the aqe-mod guard that refuses destructive operations on AQE learning data

9 skills · 9 commands · 11 agents · 1 MCP server

ruvnet/ruflo 81,589

headroom

claude

Headroom startup hooks for Claude Code and GitHub Copilot CLI.

hooks

headroomlabs-ai/headroom 79,508

mempalace

claudecodex

AI memory system — mine projects and conversations into a searchable palace. 47 MCP tools, auto-save hooks, guided setup.

3 skills · 6 commands · 1 MCP server · hooks

MemPalace/mempalace 60,536

Maintain observe?

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 plugins14 plugins