Workspace

Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...

Security Scan

The skill's stated purpose (self-reflection / memory) roughly matches its SKILL.md, but the package contains a large, fully-featured codebase and installation/automation scripts that are not declared or justified by the simple description — several mismatches raise safety and privilege concerns.

Purpose & Capability

The SKILL.md describes a self-improving memory/heartbeat system which legitimately needs to read/write a small set of workspace files. However the bundle includes hundreds of code files (searchers, scrapers, installers, system-permission scripts, skill-publisher/evolver tools, identity utilities, aggressive optimization/plagiarism-reduction tooling, get-pip, install scripts, etc.). The registry metadata said 'instruction-only' / 'no install spec' and 'no required env vars', yet the repository contains modules that expect API keys, perform web scraping, install packages, and modify skills. The presence of a large research+automation suite (slug 'bridge-research-suite') is disproportionate to the simple 'Self-Improving' description and suggests either a bundled platform or accidental mismatch.

Instruction Scope

SKILL.md instructs the agent to create and use ~/self-improving/, to read AGENTS.md, SOUL.md, USER.md and daily memory files, and to run setup.md when missing. Those operations require file-system read/write and potentially running setup scripts. The repo contains many setup/install scripts (setup_*.sh, install_system_packages.sh, install_skills.py, extract-skill.sh) and automation hooks that the SKILL.md implies could be executed; the instructions also encourage proactive background actions and pushing changes — that gives the agent broad discretion to modify files and run installers beyond the narrow 'memory' scope.

Install Mechanism

The skill is claimed to be instruction-only (no install spec) but the package includes many executable scripts, install helpers (get-pip.py, setup_*.sh, install_skills.py), and archives. Because there is no declared install mechanism, these code files might not normally run — but SKILL.md's 'run setup.md' guidance and included scripts provide a pathway to execute arbitrary installs. The repository does not point to trusted external release hosts for binary downloads; local scripts could install packages or change system state if executed.

Credentials

Registry metadata claims 'no credentials required' yet code contains components that clearly expect credentials (e.g., academic_paper_searcher.Aca demicPaperSearcher requires a ScraperAPI api_key), network access for scrapers, and identity modules (verified-agent-identity) that handle keys and signatures. SKILL.md lists configPaths like ~/self-improving/ (file-system access) and optional AGENTS.md/SOUL.md which lets the skill read and write personal workspace files. The set of requested/implicit accesses (filesystem, network, potential API keys, ability to install packages) is broader than the declared zero-env-vars, which is an incoherence and privacy/credential risk.

Persistence & Privilege

always:false (good), and default platform autonomy is allowed. However included files express 'auto_skill_override' and 'skill publisher' / 'auto_solidify_no_ask' behavior (e.g., absolute_silence_v14.json and capability-evolver/publisher code) that would autonomously solidify or publish skills without prompting. The package contains scripts for installing skills, modifying permissions, and self-repair — combined with autonomous invocation this could persist or escalate privileges. The SKILL.md itself recommends proactive background work and writing to workspace files, which grants ongoing presence in the user's filesystem.

What to consider before installing

This package contains an instruction-only SKILL.md that makes sense for a self-improving memory, but the repository also includes hundreds of scripts and automation tools that can install packages, run scrapers, change permissions, modify or publish skills, and read/write user files. Before installing or enabling this skill: 1) Do not run it with autonomous execution turned on; require explicit confirmation for any actions that write files or run scripts. 2) Inspect setup.md and any install_*.sh / install_skills.py / extract-skill.sh before allowing them to run — prefer to run these in an isolated sandbox or ephemeral VM. 3) Do not supply any API keys or credentials unless you audit exactly which component needs them (e.g., ScraperAPI) and why. 4) Search the repo for phrases like "auto_solidify", "auto_skill_override", "install_skills", "install_system_packages.sh", "get-pip.py", and "gui_permission" and review those files carefully — they indicate the skill can change system state and autonomously persist. 5) If you want only the small self-improving memory behavior, ask the author for a minimal variant that only reads/writes a narrow, user-approved directory and removes any installers, scrapers, skill-publisher, or identity tooling. 6) If you proceed, run first in an isolated environment with network disabled (or strictly limited) and no access to your real credentials or important files.

When to Use

User corrects you or points out mistakes. You complete significant work and want to evaluate the outcome. You notice something in your own output that could be better. Knowledge should compound over time without manual maintenance.

Architecture

Memory lives in ~/self-improving/ with tiered structure. If ~/self-improving/ does not exist, run setup.md. Workspace setup should add the standard self-improving steering to the workspace AGENTS, SOUL, and HEARTBEAT.md files, with recurring maintenance routed through heartbeat-rules.md.

~/self-improving/
├── memory.md          # HOT: ≤100 lines, always loaded
├── index.md           # Topic index with line counts
├── heartbeat-state.md # Heartbeat state: last run, reviewed change, action notes
├── projects/          # Per-project learnings
├── domains/           # Domain-specific (code, writing, comms)
├── archive/           # COLD: decayed patterns
└── corrections.md     # Last 50 corrections log

Quick Reference

Topic	File
Setup guide	setup.md
Heartbeat state template	heartbeat-state.md
Memory template	memory-template.md
Workspace heartbeat snippet	HEARTBEAT.md
Heartbeat rules	heartbeat-rules.md
Learning mechanics	learning.md
Security boundaries	boundaries.md
Scaling rules	scaling.md
Memory operations	operations.md
Self-reflection log	reflections.md
OpenClaw HEARTBEAT seed	openclaw-heartbeat.md

Requirements

• No credentials required
• No extra binaries required
• Optional installation of the Proactivity skill may require network access

Learning Signals

Log automatically when you notice these patterns:

Corrections → add to corrections.md, evaluate for memory.md:

• "No, that's not right..."
• "Actually, it should be..."
• "You're wrong about..."
• "I prefer X, not Y"
• "Remember that I always..."
• "I told you before..."
• "Stop doing X"
• "Why do you keep..."

Preference signals → add to memory.md if explicit:

• "I like when you..."
• "Always do X for me"
• "Never do Y"
• "My style is..."
• "For [project], use..."

Pattern candidates → track, promote after 3x:

• Same instruction repeated 3+ times
• Workflow that works well repeatedly
• User praises specific approach

Ignore (don't log):

• One-time instructions ("do X now")
• Context-specific ("in this file...")
• Hypotheticals ("what if...")

Self-Reflection

After completing significant work, pause and evaluate:

1. Did it meet expectations? — Compare outcome vs intent
2. What could be better? — Identify improvements for next time
3. Is this a pattern? — If yes, log to corrections.md

When to self-reflect:

• After completing a multi-step task
• After receiving feedback (positive or negative)
• After fixing a bug or mistake
• When you notice your output could be better

Log format:

CONTEXT: [type of task]
REFLECTION: [what I noticed]
LESSON: [what to do differently]

Example:

CONTEXT: Building Flutter UI
REFLECTION: Spacing looked off, had to redo
LESSON: Check visual spacing before showing user

Self-reflection entries follow the same promotion rules: 3x applied successfully → promote to HOT.

Quick Queries

User says	Action
"What do you know about X?"	Search all tiers for X
"What have you learned?"	Show last 10 from corrections.md
"Show my patterns"	List memory.md (HOT)
"Show [project] patterns"	Load projects/{name}.md
"What's in warm storage?"	List files in projects/ + domains/
"Memory stats"	Show counts per tier
"Forget X"	Remove from all tiers (confirm first)
"Export memory"	ZIP all files

Memory Stats

On "memory stats" request, report:

📊 Self-Improving Memory

HOT (always loaded):
  memory.md: X entries

WARM (load on demand):
  projects/: X files
  domains/: X files

COLD (archived):
  archive/: X files

Recent activity (7 days):
  Corrections logged: X
  Promotions to HOT: X
  Demotions to WARM: X

Common Traps

Trap	Why It Fails	Better Move
Learning from silence	Creates false rules	Wait for explicit correction or repeated evidence
Promoting too fast	Pollutes HOT memory	Keep new lessons tentative until repeated
Reading every namespace	Wastes context	Load only HOT plus the smallest matching files
Compaction by deletion	Loses trust and history	Merge, summarize, or demote instead

Core Rules

1. Learn from Corrections and Self-Reflection

• Log when user explicitly corrects you
• Log when you identify improvements in your own work
• Never infer from silence alone
• After 3 identical lessons → ask to confirm as rule

2. Tiered Storage

Tier	Location	Size Limit	Behavior
HOT	memory.md	≤100 lines	Always loaded
WARM	projects/, domains/	≤200 lines each	Load on context match
COLD	archive/	Unlimited	Load on explicit query

3. Automatic Promotion/Demotion

• Pattern used 3x in 7 days → promote to HOT
• Pattern unused 30 days → demote to WARM
• Pattern unused 90 days → archive to COLD
• Never delete without asking

4. Namespace Isolation

• Project patterns stay in projects/{name}.md
• Global preferences in HOT tier (memory.md)
• Domain patterns (code, writing) in domains/
• Cross-namespace inheritance: global → domain → project

5. Conflict Resolution

When patterns contradict:

1. Most specific wins (project > domain > global)
2. Most recent wins (same level)
3. If ambiguous → ask user

6. Compaction

When file exceeds limit:

1. Merge similar corrections into single rule
2. Archive unused patterns
3. Summarize verbose entries
4. Never lose confirmed preferences

7. Transparency

• Every action from memory → cite source: "Using X (from projects/foo.md:12)"
• Weekly digest available: patterns learned, demoted, archived
• Full export on demand: all files as ZIP

8. Security Boundaries

See boundaries.md — never store credentials, health data, third-party info.

9. Graceful Degradation

If context limit hit:

1. Load only memory.md (HOT)
2. Load relevant namespace on demand
3. Never fail silently — tell user what's not loaded

Scope

This skill ONLY:

• Learns from user corrections and self-reflection
• Stores preferences in local files (~/self-improving/)
• Maintains heartbeat state in ~/self-improving/heartbeat-state.md when the workspace integrates heartbeat
• Reads its own memory files on activation

This skill NEVER:

• Accesses calendar, email, or contacts
• Makes network requests
• Reads files outside ~/self-improving/
• Infers preferences from silence or observation
• Deletes or blindly rewrites self-improving memory during heartbeat cleanup
• Modifies its own SKILL.md

Data Storage

Local state lives in ~/self-improving/:

• memory.md for HOT rules and confirmed preferences
• corrections.md for explicit corrections and reusable lessons
• projects/ and domains/ for scoped patterns
• archive/ for decayed or inactive patterns
• heartbeat-state.md for recurring maintenance markers

Related Skills

Install with clawhub install <slug> if user confirms:

• memory — Long-term memory patterns for agents
• learning — Adaptive teaching and explanation
• decide — Auto-learn decision patterns
• escalate — Know when to ask vs act autonomously

Feedback

• If useful: clawhub star self-improving
• Stay updated: clawhub sync