Continuous Improvement
- 6 repo stars
- Updated August 4, 2026
- naimkatiman/continuous-improvement
Persistent-memory and runtime-discipline layer for Claude Code: remembers corrections and turns fixes into reusable instincts.
About
A persistent-memory and runtime-discipline layer for Claude Code that remembers prior corrections and grounds edits in real facts before they land. Its Mulahazah engine turns each fix into a reusable instinct so lessons apply automatically. Ships 25 skills, instinct-aware hooks, an MCP recall toolset, and a transcript-linting GitHub Action.
- Persistent corrective memory
- Mulahazah instinct engine
- 7 Laws of agent discipline
- 25 skills + hooks + MCP toolset
Continuous Improvement by the numbers
- Data as of Aug 5, 2026 (Skillselion catalog sync)
/plugin marketplace add naimkatiman/continuous-improvement/plugin install continuous-improvement@continuous-improvementAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| repo stars | ★ 6 |
|---|---|
| Last updated | August 4, 2026 |
| Repository | naimkatiman/continuous-improvement ↗ |
What it does
Persistent-memory and runtime-discipline layer for Claude Code: remembers corrections and turns fixes into reusable instincts.
README.md
Continuous Improvement Plugin Bundle (Claude Code)
This directory is generated by npm run build.
Do not edit files here manually; edit the source files in the repo root and rebuild.
Included surfaces:
.claude-plugin/plugin.json.claude-plugin/marketplace.jsonskills/commands/agents/—code-reviewer,security-auditor,test-engineerpersonas (auto-discovered Claude Code subagents; pattern from addy/agent-skills)hooks/bin/mcp-server.mjsbin/observe.mjsbin/backfill.mjslib/plugin-metadata.mjslib/observe-event.mjsinstinct-packs/templates/planning-with-files/
Required vs Optional companions
The proceed-with-the-recommendation skill (bundled here) routes each
recommendation it walks to a specialist skill. Some specialists ship in
this bundle; the rest live in external plugins. Every external routing
target has a concretely-worded inline fallback in the orchestrator skill,
so the plugin works on a clean install with no other plugins present —
the trade-off is fallback quality vs dedicated-skill quality.
Bundled with this plugin (no extra install needed):
continuous-improvement— core 7 Laws skillproceed-with-the-recommendation— the orchestrator itselfralph— autonomous PRD-scale looptdd-workflow— RED/GREEN/REFACTOR + 80% coverage gateworkspace-surface-audit— environment + capability audit- Tier-1/Tier-2 enforcement skills (
gateguard,verification-loop,safety-guard,token-budget-advisor,strategic-compact,wild-risa-balance)
Optional companions the orchestrator routes to (install separately if you want the dedicated skill instead of the inline fallback):
| Source plugin | Routing targets | Notes |
|---|---|---|
obra/superpowers |
superpowers:brainstorming, :writing-plans, :systematic-debugging, :test-driven-development, :requesting-code-review, :verification-before-completion, :dispatching-parallel-agents, :finishing-a-development-branch, simplify, schedule, loop, update-config |
Largest external surface; covers planning, debugging, parallel dispatch, branch-finish |
| Claude Code host (Anthropic built-in) | security-review |
Ships with the host CLI itself, not a separate plugin install |
code-review plugin |
code-review |
Diff walk + severity tagging |
documentation-lookup agent / context7 MCP |
documentation-lookup |
Library docs lookup with citations |
frontend-design plugin |
frontend-design:frontend-design |
Distinctive UI generation |
commit-commands plugin |
commit-commands:commit, :commit-push-pr |
Conventional-commit + push-PR helpers |
Full per-target audit (license status, bundled flag, inline fallback,
risk-if-absent) lives in docs/audits/2026-05-06-routing-target-gap.md.
Per-target inline fallbacks are spelled out in the orchestrator skill at
skills/proceed-with-the-recommendation/SKILL.md § "Routing Table
(with Inline Fallbacks)".
Related skills
Grill MeHave an agent relentlessly pressure-test their plan or design until every decision branch is resolved and shared understanding is reached.691k196kGrill With DocsPressure-test architecture decisions against their live codebase and living documentation before writing production code.585k196k
Lark CalendarLet their coding agents read, create, and manage events inside Lark (Feishu) Calendar directly from natural language instructions.493kLark ImLet their coding agent send instant notifications, post updates, and receive commands directly inside Lark (Feishu) workspaces.493kLark ContactLet their AI coding agents read and write contacts, send messages, and manage user data inside Lark (Feishu).493kLark AttendanceAutomatically fetch and report daily attendance data from Lark (Feishu) into their own tools or agents.493k