Claude Reflect
Captures your mid-session corrections and syncs them into CLAUDE.md as permanent memory.
Test report
- Verdict
- Tested · Works
- Score
- Tested
- Jul 20, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · c821759
Cloned the repo and ran it for real. detect_patterns correctly captured all 4 corrections (e.g. "no, use gpt-5.1 not gpt-5" conf 0.85; "remember: always use venv" 0.90; Spanish "no, usa Python en vez de Node" 0.80 via regex) and correctly rejected "What file should I edit next?" and "No problem, thanks!". Full test suite: 222 passed in 0.8s (README undersells at "160"). Clean: no network/exfil, only subprocess call is the local `claude` CLI for the optional semantic layer. Frontmatter has name+description; all referenced files (commands, scripts, assets/reflect-demo.jpg, hooks.json using ${CLAUDE_PLUGIN_ROOT}) exist; no hardcoded machine paths. Real friction: it's a plugin needing marketplace-add + install + a Claude Code restart, not a drop-in SKILL.md, and the final CLAUDE.md quality still depends on the interactive /reflect review I did not drive end-to-end. Minor doc drift: README badge v2.6.0 vs plugin.json v3.1.0.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 5/5
- Output vs. baseline 8/10
- Docs & honesty 4/5
What Claude Reflect does
A Claude Code plugin that hooks every prompt, detects correction patterns ("no, use X not Y", "remember:") and queues them, then /reflect reviews and writes them into CLAUDE.md/AGENTS.md so Claude stops repeating mistakes. Triggers when you correct Claude or ask it to remember something for future sessions.
How to install Claude Reflect
claude plugin marketplace add bayramannakov/claude-reflect && claude plugin install claude-reflect@claude-reflect-marketplace (then restart Claude Code)
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Claude Reflect
-
/claude-reflectCaptures your mid-session corrections and syncs them into CLAUDE.md as permanent memory.
It also activates on plain-language prompts like these:
-
Remember this correction for future sessions -
Run reflect to update CLAUDE.md with today's learnings -
Save this preference so I don't repeat myself
Frequently asked questions
- Is the Claude Reflect skill free?
- Yes. The skill itself is free from BayramAnnakov/claude-reflect. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Claude Reflect work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 20, 2026. Verdict: Tested · Works. Cloned the repo and ran it for real. detect_patterns correctly captured all 4 corrections (e.g. "no, use gpt-5.1 not gpt-5" conf 0.85; "remember: always use venv" 0.90; Spanish "no, usa Python en vez de Node" 0.80 via regex) and correctly rejected "What file should I edit next?" and "No problem, thanks!". Full test suite: 222 passed in 0.8s (README undersells at "160"). Clean: no network/exfil, only subprocess call is the local `claude` CLI for the optional semantic layer. Frontmatter has name+description; all referenced files (commands, scripts, assets/reflect-demo.jpg, hooks.json using ${CLAUDE_PLUGIN_ROOT}) exist; no hardcoded machine paths. Real friction: it's a plugin needing marketplace-add + install + a Claude Code restart, not a drop-in SKILL.md, and the final CLAUDE.md quality still depends on the interactive /reflect review I did not drive end-to-end. Minor doc drift: README badge v2.6.0 vs plugin.json v3.1.0.
- What is the Claude Reflect SkillProof Score?
- 8.8/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 4/5.
- How do I install Claude Reflect?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Claude Reflect with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.