The Agent Skills Directory

[COMMAND_EXECUTION]: The skill and its associated shell script (scripts/generate-report.sh) execute several commands to interact with GitHub and messaging platforms:
Uses gh search issues and gh search prs to fetch repository data.
Uses gh issue view and gh issue comment to manage issue follow-ups.
Uses clawdbot message send to exfiltrate generated reports to a specific Discord channel (ID: 1465549185632702591).
Utilizes standard utilities like jq, sed, and tr for JSON processing and text sanitization.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) because it ingests and processes untrusted content from GitHub issues and pull requests.
Ingestion points: Data is fetched via gh CLI commands in both SKILL.md and generate-report.sh (e.g., issue titles, bodies, and comments from the alibaba/higress repository).
Boundary markers: The instructions do not define explicit delimiters or warnings to the AI agent to ignore instructions embedded within the fetched GitHub content.
Capability inventory: The agent has the ability to read/write local files (in /root/clawd/memory/) and perform network operations via the gh and clawdbot tools.
Sanitization: The shell script performs minimal sanitization (stripping backticks and truncating length), but does not prevent the AI from interpreting malicious instructions contained within the text. This is a common risk factor for summarization agents.

higress-daily-report