install structsd

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capa...

Security Scan

This is an instruction-only self-improvement skill that consistently instructs the agent to log learnings and errors to repo files; its requirements and actions align with its stated purpose, but it will persist conversational/content data to workspace files so you should be aware of privacy and repo-change implications.

Purpose & Capability

Name/description (self-improvement / skill-authoring guidance) match the instructions: the SKILL.md tells the agent to log errors, corrections, and feature requests and to promote broad learnings into project docs. No unrelated binaries, env vars, or installs are requested.

Instruction Scope

The instructions explicitly tell the agent to write logs and promote content to files such as .learnings/ERRORS.md, .learnings/LEARNINGS.md, CLAUDE.md, AGENTS.md, TOOLS.md, and .github/copilot-instructions.md. That means conversational or operational data will be persisted in the workspace and may alter repository files; sensitive user data could be recorded unintentionally. The instructions do not provide sanitization or redaction guidance.

Install Mechanism

No install spec and no code files — instruction-only skill. Nothing is downloaded or written by an installer.

Credentials

No environment variables, credentials, or config paths are requested. However, the skill's behavior will create/modify repository files, which is a form of persistence rather than credential access.

Persistence & Privilege

always:false (normal). The skill will nonetheless persist data to workspace files and recommend promoting findings to project-wide docs; this is expected for its purpose but increases the blast radius of any sensitive content the agent logs. It does not request elevated platform/global permissions or modify other skills' configurations.

Assessment

This skill is coherent for self-improvement: it tells the agent to log learnings and errors into repository files. Before enabling it, decide whether you want the agent to persist conversational content in the workspace. Consider these mitigations: add .learnings and any auto-generated files to .gitignore or a private-only area; update the instructions to redact or omit secrets before logging; restrict the agent's write access to a sandbox folder; review any logs regularly and remove sensitive entries; and explicitly test the skill in a non-sensitive workspace first. If you do not want persistent logs or automatic promotion to shared docs, do not enable this skill or modify its instructions to limit what gets recorded.

Self-Improvement Skill

Log learnings and errors to markdown files for continuous improvement. Coding agents can later process these into fixes, and important learnings get promoted to project memory.

Quick Reference

Situation	Action
Command/operation fails	Log to .learnings/ERRORS.md
User corrects you	Log to .learnings/LEARNINGS.md with category correction
User wants missing feature	Log to .learnings/FEATURE_REQUESTS.md
API/external tool fails	Log to .learnings/ERRORS.md with integration details
Knowledge was outdated	Log to .learnings/LEARNINGS.md with category knowledge_gap
Found better approach	Log to .learnings/LEARNINGS.md with category best_practice
Simplify/Harden recurring patterns	Log/update .learnings/LEARNINGS.md with Source: simplify-and-harden and a stable Pattern-Key
Similar to existing entry	Link with **See Also**, consider priority bump
Broadly applicable learning	Promote to CLAUDE.md, AGENTS.md, and/or .github/copilot-instructions.md
Workflow improvements	Promote to AGENTS.md (OpenClaw workspace)
Tool gotchas	Promote to TOOLS.md (OpenClaw workspace)
Behavioral patterns	Promote to SOUL.md (OpenClaw workspace)