MCPInstall

reflect

Analyze the current session and propose improvements to skills. Proactively invoke this skill when you notice user corrections after skill usage, or at the end of skill-heavy sessions. Also use when user says "reflect", "improve skill", or "learn from this".

Installs
45
Last 24h
—
Trending rank
—
Security
Not audited

Install

Works with every agent the skills CLI supports.

npx skills add https://github.com/llama-farm/llamafarm --skill reflect
  1. 01

    Install

    Run the command in your project. The CLI asks which agents to install for.

  2. 02

    Describe the task

    Ask your agent for something this skill covers. It loads reflect automatically when relevant.

  3. 03

    Review the result

    The agent follows the SKILL.md steps, scripts and references bundled with the skill.

SKILL.md

The instructions your agent reads when the skill is triggered.

Reflect Skill

Analyze the current session and propose improvements to skills based on what worked, what didn't, and edge cases discovered.

Trigger

Run /reflect or /reflect [skill-name] after a session where you used a skill.

Additional commands:

  • /reflect on - Enable automatic end-of-session reflection
  • /reflect off - Disable automatic reflection
  • /reflect status - Check if auto-reflect is enabled

Workflow

Step 1: Identify the Skill

If skill name not provided, ask:

Which skill should I analyze this session for?
- frontend-design
- code-reviewer
- [other]

Step 2: Analyze the Conversation

Look for these signals in the current conversation:

Corrections (HIGH confidence):

  • User said "no", "not like that", "I meant..."
  • User explicitly corrected output
  • User asked for changes immediately after generation

Successes (MEDIUM confidence):

  • User said "perfect", "great", "yes", "exactly"
  • User accepted output without modification
  • User built on top of the output

Edge Cases (MEDIUM confidence):

  • Questions the skill didn't anticipate
  • Scenarios requiring workarounds
  • Features user asked for that weren't covered

Preferences (accumulate over sessions):

  • Repeated patterns in user choices
  • Style preferences shown implicitly
  • Tool/framework preferences

Step 3: Propose Changes

Present findings using accessible colors (WCAG AA 4.5:1 contrast ratio):

┌─ Skill Reflection: [skill-name] ───────────────────┐
│                                                    │
│ Signals: X corrections, Y successes                │
│                                                    │
│ Proposed changes:                                  │
│                                                    │
│ 🔴 [HIGH] + Add constraint: "[specific constraint]"│
│ 🟡 [MED]  + Add preference: "[specific preference]"│
│ 🔵 [LOW]  ~ Note for review: "[observation]"       │
│                                                    │
│ Commit: "[skill]: [summary of changes]"            │
│                                                    │
└────────────────────────────────────────────────────┘

Apply these changes? [Y/n] or describe tweaks

Accessible Color Palette (use ANSI codes in terminal output):

  • HIGH: \033[1;31m (bold red #FF6B6B - 4.5:1 on dark)
  • MED: \033[1;33m (bold yellow #FFE066 - 4.8:1 on dark)
  • LOW: \033[1;36m (bold cyan #6BC5FF - 4.6:1 on dark)
  • Reset: \033[0m

Avoid: pure red (#FF0000) on black, green on red (colorblind users)

User Response Options:

  • Y – Apply changes, commit, and push
  • n – Skip this update
  • Or describe any tweaks to the proposed changes

Step 4: If Approved

  1. Read the current skill file from .claude/skills/[skill-name]/SKILL.md
  2. Apply the changes using the Edit tool
  3. Run git commands:
    cd .claude/skills
    git add [skill-name]/SKILL.md
    git commit -m "[skill]: [change summary]"
    git push origin main
    
  4. Confirm: "Skill updated and pushed to GitHub"

Step 5: If Declined

Ask: "Would you like to save these observations for later review?"

If yes, append to .claude/skills/[skill-name]/OBSERVATIONS.md

Toggle Commands

/reflect on

Enable automatic end-of-session reflection:

  1. Create/update ~/.claude/reflect-skill-state.json with {"enabled": true, "updatedAt": "[timestamp]"}
  2. Confirm: "Auto-reflect enabled. Sessions will be analyzed automatically when you stop."

/reflect off

Disable automatic reflection:

  1. Update ~/.claude/reflect-skill-state.json with {"enabled": false, "updatedAt": "[timestamp]"}
  2. Confirm: "Auto-reflect disabled. Use /reflect manually to analyze sessions."

/reflect status

Check current status:

  1. Read ~/.claude/reflect-skill-state.json
  2. Report: "Auto-reflect is [enabled/disabled]" with last updated timestamp

Note: The state file is saved in the global Claude user directory (~/.claude/) so it persists across plugin upgrades.

Example Session

User runs /reflect frontend-design after a UI session:

┌─ Skill Reflection: frontend-design ───────────────┐
│                                                   │
│ Signals: 2 corrections, 3 successes               │
│                                                   │
│ Proposed changes:                                 │
│                                                   │
│ 🔴 [HIGH] + Constraints/NEVER:                    │
│    "Use gradients unless explicitly requested"    │
│                                                   │
│ 🔴 [HIGH] + Color & Theme:                        │
│    "Dark backgrounds: use #000, not #1a1a1a"      │
│                                                   │
│ 🟡 [MED] + Layout:                                │
│    "Prefer CSS Grid for card layouts"             │
│                                                   │
│ Commit: "frontend-design: no gradients, #000 dark"│
│                                                   │
└───────────────────────────────────────────────────┘

Apply these changes? [Y/n] or describe tweaks

Git Integration

This skill has permission to:

  • Read skill files from .claude/skills/
  • Edit skill files (with user approval)
  • Run git add, git commit, git push in the skills directory

The skills repo should be initialized at .claude/skills with a remote origin.

Important Notes

  • Always show the exact changes before applying
  • Never modify skills without explicit user approval
  • Commit messages should be concise and descriptive
  • Push only after successful commit

FAQ

What does the reflect skill do?

Analyze the current session and propose improvements to skills. Proactively invoke this skill when you notice user corrections after skill usage, or at the end of skill-heavy sessions. Also use when user says "reflect", "improve skill", or "learn from this".

How do I install reflect?

Run npx skills add https://github.com/llama-farm/llamafarm --skill reflect in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/reflect/ (personal) or .claude/skills/reflect/ (project).

Which AI agents can use reflect?

Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.

Is reflect safe to install?

No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.

Who maintains reflect?

It is published by llama-farm in llama-farm/llamafarm. It has 45 recorded installs.

More from llama-farm/llamafarm

View all →

code-review

llama-farm/llamafarm

Comprehensive code review for diffs. Analyzes changed code for security vulnerabilities, anti-patterns, and quality issues. Auto-detects domain (frontend/backend) from file paths.

commit-push-pr

llama-farm/llamafarm

Open the skill to read its SKILL.md instructions.

electron-skills

llama-farm/llamafarm

Open the skill to read its SKILL.md instructions.

85 installs

typescript-skills

llama-farm/llamafarm

Open the skill to read its SKILL.md instructions.

python-skills

llama-farm/llamafarm

Open the skill to read its SKILL.md instructions.

68 installs

react-skills

llama-farm/llamafarm

React 18 patterns for LlamaFarm Designer. Covers components, hooks, TanStack Query, and testing.

Popular education & science skills

View category →

teach

mattpocock/skills

Teach the user a new skill or concept, within this workspace.

Education & Science874K installs

scaffold-exercises

mattpocock/skills

Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use when user wants to scaffold exercises, create exercise stubs, or set up a new course section.

Education & Science488.2K installs

paper-context-resolver

lllllllama/rigorpilot-skills

Rigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical gap and the task is to resolve a specific paper detail such as dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption from primary paper sources while recording conflicts. Do not use for general paper summary, repo scanning, environment setup, command execution, title-only paper lookup, or replacing README guidance by default.

Education & Science478.3K installs

caveman-learn

juliusbrussee/caveman

Act on a Caveman learn report - review the ranked token sinks, apply cost-lowering fixes with per-edit consent, and report what those fixes returned. Use when asked to lower an agent's token cost, what caveman has saved, to trim a heavy CLAUDE.md, or to offload re-pasted context into cavemem.

Education & Science107.9K installs

gsap-plugins

greensock/gsap-skills

Official GSAP skill for GSAP plugins — registration, ScrollToPlugin, ScrollSmoother, Flip, Draggable, Inertia, Observer, SplitText, ScrambleText, SVG and physics plugins, CustomEase, EasePack, CustomWiggle, CustomBounce, GSDevTools. Use when the user asks about a GSAP plugin, scroll-to, flip animations, draggable, SVG drawing, or plugin registration.

Education & Science67.9K installs

teach-impeccable

pbakaus/impeccable

DEPRECATED: Use /impeccable teach instead. This command has been folded into the impeccable skill.