{"id":49535489,"url":"https://github.com/naimkatiman/continuous-improvement","last_synced_at":"2026-05-02T10:03:37.213Z","repository":{"id":349080674,"uuid":"1200990495","full_name":"naimkatiman/continuous-improvement","owner":"naimkatiman","description":"Install 7-law discipline + instinct-based learning (Mulahazah) into Claude Code — hooks observe every session, instincts strengthen or decay with confidence scoring","archived":false,"fork":false,"pushed_at":"2026-04-27T02:49:12.000Z","size":3772,"stargazers_count":3,"open_issues_count":23,"forks_count":3,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-04-27T03:30:13.134Z","etag":null,"topics":["ai-agents","ai-discipline","anthropic","claude-code","claude-code-marketplace","claude-code-plugin","codex","coding-agents","continuous-improvement","cursor","developer-tools","github-action","hooks","instinct-learning","llm","mcp","mulahazah","prompt-engineering","system-prompt"],"latest_commit_sha":null,"homepage":null,"language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/naimkatiman.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-04T04:20:45.000Z","updated_at":"2026-04-27T02:49:14.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/naimkatiman/continuous-improvement","commit_stats":null,"previous_names":["naimkatiman/continuous-improve-skill","naimkatiman/continuous-improvement"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/naimkatiman/continuous-improvement","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/naimkatiman%2Fcontinuous-improvement","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/naimkatiman%2Fcontinuous-improvement/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/naimkatiman%2Fcontinuous-improvement/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/naimkatiman%2Fcontinuous-improvement/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/naimkatiman","download_url":"https://codeload.github.com/naimkatiman/continuous-improvement/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/naimkatiman%2Fcontinuous-improvement/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32530178,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-02T01:12:54.858Z","status":"online","status_checked_at":"2026-05-02T02:00:05.923Z","response_time":132,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-agents","ai-discipline","anthropic","claude-code","claude-code-marketplace","claude-code-plugin","codex","coding-agents","continuous-improvement","cursor","developer-tools","github-action","hooks","instinct-learning","llm","mcp","mulahazah","prompt-engineering","system-prompt"],"created_at":"2026-05-02T10:03:30.729Z","updated_at":"2026-05-02T10:03:37.206Z","avatar_url":"https://github.com/naimkatiman.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e\n  \u003cimg src=\"assets/combined.gif\" alt=\"Before vs After — The 7 Laws of AI Agent Discipline\" width=\"700\" /\u003e\n\u003c/p\u003e\n\n\u003ch1 align=\"center\"\u003eThe 7 Laws of AI Agent Discipline\u003c/h1\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cb\u003eStop your Claude Code agent from skipping steps, guessing, and declaring \"done\" without verifying.\u003c/b\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://www.npmjs.com/package/continuous-improvement\"\u003e\u003cimg src=\"https://img.shields.io/npm/v/continuous-improvement\" alt=\"npm\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://docs.anthropic.com/en/docs/claude-code\"\u003e\u003cimg src=\"https://img.shields.io/badge/Claude%20Code-skill-blueviolet\" alt=\"Claude Code\"\u003e\u003c/a\u003e\n  \u003ca href=\"LICENSE\"\u003e\u003cimg src=\"https://img.shields.io/badge/license-MIT-green\" alt=\"license\"\u003e\u003c/a\u003e\n  \u003ca href=\"test/\"\u003e\u003cimg src=\"https://img.shields.io/badge/tests-passing-brightgreen\" alt=\"tests\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cb\u003eNew here?\u003c/b\u003e → \u003ca href=\"QUICKSTART.md\"\u003eQUICKSTART.md\u003c/a\u003e (2 minutes)\n\u003c/p\u003e\n\n---\n\n## Install\n\n### Option A — Claude Code plugin (recommended, 2 commands)\n\nRun inside the Claude Code CLI:\n\n```bash\n/plugin marketplace add naimkatiman/continuous-improvement\n/plugin install continuous-improvement@continuous-improvement\n```\n\nUpdate later with `/plugin marketplace update continuous-improvement`. Browse the other 8 PM plugins with `/plugin` (interactive picker) or see the [full list below](#plugin-marketplace-plugins).\n\n### Option B — npm CLI installer\n\nRequires Node 18/20/22 and `bash` (Git Bash or WSL on Windows).\n\n```bash\n# Beginner — 7 Laws skill, hooks, slash commands\nnpx continuous-improvement install\n\n# Expert — adds MCP server (12 tools) + session hooks\nnpx continuous-improvement install --mode expert\n\n# Optional starter packs\nnpx continuous-improvement install --pack react|python|go\n```\n\nVerify either path with `/discipline` or `/dashboard` in Claude Code.\n\n---\n\n## The 7 Laws\n\n| # | Law | Without it, agents... |\n|---|-----|----------------------|\n| 1 | **Research Before Executing** | reinvent what already exists |\n| 2 | **Plan Is Sacred** | scope-creep and overbuild |\n| 3 | **One Thing at a Time** | stack untested changes |\n| 4 | **Verify Before Reporting** | lie about being \"done\" |\n| 5 | **Reflect After Sessions** | repeat the same failures |\n| 6 | **Iterate One Change** | debug 5 changes at once |\n| 7 | **Learn From Every Session** | lose knowledge when context ends |\n\n```\nResearch -\u003e Plan -\u003e Execute (one thing) -\u003e Verify -\u003e Reflect -\u003e Learn -\u003e Iterate\n```\n\n\u003cp align=\"center\"\u003e\n  \u003cimg src=\"assets/diagram-7-laws-loop.jpg\" alt=\"The 7 Laws of AI Agent Discipline — circular workflow loop\" width=\"820\" /\u003e\n\u003c/p\u003e\n\nFull spec, reflection-block format, and anti-examples: [SKILL.md](SKILL.md).\n\n---\n\n## Mulahazah: Auto-Leveling Learning\n\nHooks capture every tool call. After ~20 observations, Claude analyzes patterns and creates **instincts** with confidence scores:\n\n- **\u003c 0.5** silent (stored, not surfaced)\n- **0.5–0.69** suggested inline when relevant\n- **0.7+** auto-applied\n- User corrections drop confidence by 0.1; unused instincts decay\n- Project-scoped, promoted to global after seen across 2+ projects\n\n\u003cp align=\"center\"\u003e\n  \u003cimg src=\"assets/diagram-mulahazah-learning.jpg\" alt=\"Mulahazah pipeline\" width=\"820\" /\u003e\n\u003c/p\u003e\n\n---\n\n## Slash Commands\n\n```\n/continuous-improvement    Reflect, analyze, show status\n/planning-with-files       Create task_plan.md, findings.md, progress.md\n/discipline                Quick reference card of the 7 Laws\n/dashboard                 Visual instinct health dashboard\n```\n\nIn expert mode, the same planning workflow is also available programmatically through the MCP tools `ci_plan_init` (initialize `task_plan.md`, `findings.md`, `progress.md` in the project root) and `ci_plan_status` (summarize their current contents).\n\n---\n\n## Bundled Skills \u0026 Plugins\n\nThe default install only deploys the core 7 Laws skill. Everything below is opt-in — pick what you want.\n\n### Featured companion — the recommended pairing for the 7 Laws\n\n⭐ **`proceed-with-the-recommendation`** is the execution arm of this repo. When you say \"do all of it\" / \"proceed with your recommendation\" / \"yes do it\", this skill walks the list top-to-bottom under the 7 Laws, routes each item to the right specialist (`superpowers:*`, `ralph`, `workspace-surface-audit`, `simplify`, `security-review`, `schedule`, `loop`), falls back to inline behavior when a specialist isn't installed, verifies per item, and **halts on `needs-approval`** instead of barreling through. If you adopt one companion alongside the core skill, adopt this one — it's purpose-built to operationalize the laws end-to-end.\n\nInstall it on its own:\n\n```bash\nSKILL=proceed-with-the-recommendation\nmkdir -p ~/.claude/skills/$SKILL\ncurl -L https://raw.githubusercontent.com/naimkatiman/continuous-improvement/main/skills/$SKILL.md \\\n  -o ~/.claude/skills/$SKILL/SKILL.md\n```\n\n…or get it (and the rest below) by installing the `continuous-improvement` plugin from the marketplace.\n\n### Tier 1 — recommended pairing for **beginner** install\n\nThese add concrete enforcement to the 7 Laws and ship in every plugin install.\n\n| Skill | What it does | Pairs with |\n|-------|--------------|------------|\n| `para-memory-files` | File-based persistent memory using PARA (Projects/Areas/Resources/Archives) for cross-session context | Law 5 + Law 7 |\n| `verification-loop` | Six-phase verification (build, types, lint, tests, security, diff) with PASS/FAIL report | Law 4 |\n| `gateguard` | PreToolUse fact-forcing gate that blocks Edit/Write/destructive Bash until concrete investigation is presented | Law 1 + Law 3 |\n| `tdd-workflow` | RED→GREEN→REFACTOR enforcement with 80%+ coverage gate across unit/integration/E2E | Law 3 + Law 4 |\n\n### Tier 2 — additional skills for **expert** install\n\nLayer on top of tier-1 for autonomous-mode safety, response-depth control, and context-window discipline.\n\n| Skill | What it does |\n|-------|--------------|\n| `safety-guard` | Three-mode runtime guard (careful/freeze/guard) that blocks destructive commands and locks edits to a directory |\n| `token-budget-advisor` | Heuristic input/output token estimator that offers 25%/50%/75%/100% depth choices before answering |\n| `strategic-compact` | PreToolUse hook that suggests `/compact` at logical phase boundaries instead of arbitrary auto-compaction |\n| `wild-risa-balance` | Decision-framing lens that pairs WILD (bold) generation with RISA (safe) execution; splits recommendation lists into pilots above a baseline so bold options stop losing to safe ones in a flat list |\n\nThe `/learn-eval` slash command also ships with the expert install: extract a session pattern, run a checklist quality gate, and decide global-vs-project save location before writing any skill file.\n\n### Always-bundled companion skills ([`skills/`](skills/))\n\nDrop-in single-file skills, copy to `~/.claude/skills/\u003cname\u003e/SKILL.md`.\n\n| Skill | What it does | Source |\n|-------|--------------|--------|\n| `ralph` | Autonomous loop that executes a PRD story-by-story with quality checks between iterations | [snarktank/ralph](https://github.com/snarktank/ralph) |\n| `superpowers` | Activates task-appropriate skills automatically (brainstorming, git-worktrees, TDD, code review, etc.) | [obra/superpowers](https://github.com/obra/superpowers) |\n| `workspace-surface-audit` | Audits the active repo, MCP servers, plugins, and env, then recommends high-value skills/workflows | continuous-improvement |\n\n### Plugin marketplace ([`plugins/`](plugins/))\n\nInstall via `/plugin marketplace add naimkatiman/continuous-improvement` then `/plugin install \u003cname\u003e@continuous-improvement` (the `-dev` suffix is only used for the local in-repo development marketplace at `plugins/continuous-improvement/.claude-plugin/marketplace.json`).\n\n| Plugin | Skills | Focus |\n|--------|--------|-------|\n| `continuous-improvement` | 4 | The 7 Laws + Mulahazah (this repo's core plugin bundle) |\n| `pm-product-discovery` | 12 | Ideation, experiments, assumption testing, feature prioritization, interview synthesis |\n| `pm-product-strategy` | 12 | Vision, lean canvas, business model, SWOT, PESTLE, Ansoff, Porter's Five Forces, monetization |\n| `pm-execution` | 15 | PRDs, OKRs, roadmaps, sprints, pre-mortems, user stories, retros, release notes |\n| `pm-market-research` | 7 | Personas, market segments, market sizing, journey maps, sentiment, competitor analysis |\n| `pm-data-analytics` | 3 | SQL query generation, cohort analysis, A/B test analysis |\n| `pm-go-to-market` | 6 | GTM strategy, growth loops, GTM motions, beachhead segments, ICP, battlecards |\n| `pm-marketing-growth` | 5 | Marketing ideas, value props, North Star metrics, product naming, positioning |\n| `pm-toolkit` | 4 | Resume review, NDA drafting, privacy policy, grammar check |\n\nPM plugins by [Paweł Huryn](https://www.productcompass.pm).\n\n---\n\n## GitHub Action: Transcript Linter\n\nLint agent behavior in CI. Detects skipped laws.\n\n```yaml\n- uses: naimkatiman/continuous-improvement@v3\n  with:\n    transcript-path: agent-log.jsonl\n    strict: true\n```\n\nCatches: writes without prior research (Law 1), too many edits without verification (Law 3), code changes without tests/builds (Law 4), too many files at once (Law 6). Run locally with `node bin/lint-transcript.mjs \u003cfile\u003e`.\n\n---\n\n## Uninstall\n\n```bash\nnpx continuous-improvement install --uninstall\n```\n\nRemoves skill, hooks, commands, MCP server. Learned instincts in `~/.claude/instincts/` are preserved — delete manually for a clean slate.\n\n---\n\n## More\n\n- [QUICKSTART.md](QUICKSTART.md) — 2-minute setup\n- [SKILL.md](SKILL.md) — full 7 Laws spec\n- [examples/](examples/) — bug fix, feature build, refactor walkthroughs\n- [CONTRIBUTING.md](CONTRIBUTING.md) — architecture, repo internals\n- [SECURITY.md](SECURITY.md)\n\nMIT.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnaimkatiman%2Fcontinuous-improvement","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnaimkatiman%2Fcontinuous-improvement","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnaimkatiman%2Fcontinuous-improvement/lists"}