Agent Skills by dalomeve
Showing 25 vetted skills indexed across 1 GitHub repositories.
gateway-token-doctor
Diagnose and fix gateway token mismatches causing 401 errors across config, service, and CLI surfaces.
instruction-anchor-guard
Persist user-specified anchors with scopes and drift checks before actions.
issue-driven-skill-selector
Collects GitHub issues, validates links, and scores fixes by impact, solvability, cost, and risk.
clawhub-web-only-publish
Publish ClawHub skills through the web dashboard without CLI login.
ren-wu-shou-wei-qi
Validate task completion with DONE_CHECKLIST and EVIDENCE artifacts.
model-route-guard
Audit openclaw.json and agent overrides to resolve model routing conflicts.
skill-orchestrator
Coordinate local skills with ClawHub patterns into unified workflows.
prepublish-privacy-scrub
Scan skill files for API keys, tokens, secrets, and personal data, then redact and verify.
evidence-link-auditor
Scan markdown files to validate internal and external evidence links with PASS/FAIL reports.
powershell-reliable
Execute PowerShell commands on Windows with error handling and checkpoint resume.
phoenix-loop
Generate validated recovery playbooks from recurring failure signatures.
openclaw-growth-pack
Automates onboarding of devices onto a platform with centralized security and remote management controls.
zh-encoding-fix
Detect and repair mojibake in Markdown, JSON, and text files with UTF-8 enforcement.
clawhub-web-publisher
Publish local OpenClaw skills to ClawHub via web dashboard with secret redaction.
path-normalizer
Normalize tilde-based file paths to absolute paths for file tools.
cn-vpn-foreign-app-link
Route OpenClaw WebSocket traffic through VPN/TUN to restore foreign app channel linking.
failure-pattern-learner
Persist failure events to daily JSONL files and load relevant lessons before retries.
weekly-self-improve-loop
Aggregates weekly memory data to surface friction patterns and blocked items.
blocker-min-input
Capture exact blocker text, attempts, and minimum unblock input from tool outputs.
memory-self-heal
Automate recovery from recurring task failures using past fixes and runtime evidence.
powershell-safe-chain
Chain PowerShell commands with Try/Catch and ErrorAction Stop.
memory-to-skill-crystallizer
Converts memory logs into reusable skills with SKILL.md frontmatter.
task-finish-contract
Enforce task completion with GOAL, PROGRESS, NEXT, and EVIDENCE artifacts.
evidence-url-verifier
Verify evidence URLs resolve with HTTP 200 and real content.