The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses curl to fetch data from the GitHub API. While it targets a well-known service (GitHub), the use of system commands for network operations is a capability that should be noted.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data (GitHub issue bodies and comments) and uses the LLM to generate summaries and action items.
Ingestion points: scripts/parse_api_response.py fetches issue bodies and comments from the GitHub API.
Boundary markers: None identified in the prompt templates; the LLM is simply instructed to read the context field and generate summaries.
Capability inventory: The skill uses curl for network requests and executes Python scripts (scripts/combine_json.py, scripts/parse_api_response.py, scripts/parse_triage_readme.py) via subprocesses.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from GitHub before it is passed to the LLM for summarization.

flutter-triage