Self-Improving Agent (ORBIT)
v1.0.11Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⭐ 0· 62·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name/description (capture learnings/errors) align with what the files and scripts do: create and maintain .learnings/*.md, inject lightweight reminders at bootstrap, detect command errors, and help extract learnings into new skills. No unrelated credentials/binaries are requested. The feature set (hooks + local scripts + extract helper) is appropriate for a self‑improvement/logging skill.
Instruction Scope
SKILL.md instructs the agent and user to create .learnings/ in project or home workspace, optionally copy/enable hooks so the skill can inject reminders at agent bootstrap, and to promote learnings into workspace files (SOUL.md, AGENTS.md, etc.). The hook implementation only injects a virtual reminder file and the Bash scripts only read CLAUDE_TOOL_OUTPUT and write to stdout; they do not exfiltrate data or modify remote endpoints. However: enabling the hook intentionally allows the skill to add context into every session via workspace injection — this is expected for the purpose but is a behavior you should consciously approve.
Install Mechanism
No automated install spec; manual install uses git clone of a GitHub repo (repository referenced in SKILL.md). There are local scripts included (activator, error-detector, extract-skill) that run locally. No downloads from untrusted hosts, no archive extraction, and scripts include reasonable safety checks. This is low-risk for an instruction-only/local helper skill.
Credentials
The skill requires no environment variables or credentials. The error-detector reads CLAUDE_TOOL_OUTPUT (an agent-provided tool output variable) which is appropriate for detecting command failures. The extract-skill script writes to a relative ./skills directory and includes checks to avoid absolute paths or '..' path segments. No secrets are requested or used.
Persistence & Privilege
always:false (not force-included). Hook installation is opt-in; if you copy and enable the provided OpenClaw hook it will inject a virtual reminder at agent bootstrap. That gives the skill the ability to influence session context by adding a short reminder file — which is coherent with its purpose but should be treated as a prompt‑injection capability you enable only for trusted skills. Scripts run with the same permissions as the agent (documented in references).
Assessment
This skill appears to be what it claims: lightweight reminders, local logging, and helpers for turning learnings into files or new skills. Before enabling it or installing hooks: 1) Inspect the hook handler and scripts (they are small and readable) to confirm you’re comfortable with them writing files under your workspace and printing reminders to sessions. 2) Only enable the OpenClaw/Claude hooks if you trust the skill source; enabling hooks lets the skill inject context into every session (intended behavior for a reminder tool but a prompt‑injection capability nonetheless). 3) Verify the installation source: SKILL.md references a GitHub repo and the embedded _meta.json ownerId differs slightly from the registry owner slug — this may be a fork or renamed package; confirm the repository and publisher before cloning. 4) Run scripts in dry-run mode (extract-skill supports --dry-run) or from a controlled workspace first. 5) If you use the extract-skill helper, note it will create files under the relative ./skills path you provide — ensure you run it from the intended directory. If any of these checks raise doubt (mismatched repository, unexpected remote endpoints, or if you don’t want session-level injections), don’t enable the hooks and use the skill manually by keeping .learnings/ files in your project.Like a lobster shell, security has layers — review code before you run it.
continuous-improvementvk978aw91sqyjvhsnzxa85xj2gx83g3h9correctionsvk978aw91sqyjvhsnzxa85xj2gx83g3h9errorsvk978aw91sqyjvhsnzxa85xj2gx83g3h9latestvk978aw91sqyjvhsnzxa85xj2gx83g3h9learningvk978aw91sqyjvhsnzxa85xj2gx83g3h9orbitvk978aw91sqyjvhsnzxa85xj2gx83g3h9pt-BRvk978aw91sqyjvhsnzxa85xj2gx83g3h9self-improvementvk978aw91sqyjvhsnzxa85xj2gx83g3h9
License
MIT-0
Free to use, modify, and redistribute. No attribution required.