The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by reading existing Pull Request or Merge Request descriptions (untrusted data) through gh pr view or glab mr view and incorporating that content into its processing flow.
Ingestion points: The agent is instructed to fetch and read the current PR body in Step 4 and the 'Forge cheatsheet' section of SKILL.md.
Boundary markers: There are no explicit instructions or delimiters used to separate the existing PR content from the agent's instructions, nor are there warnings to ignore embedded instructions within the fetched data.
Capability inventory: The skill has the capability to execute shell commands via git, gh, and glab, as well as perform file writes to /tmp.
Sanitization: There is no evidence of sanitization, filtering, or escaping of the fetched PR body before the agent processes and rewrites it.
[COMMAND_EXECUTION]: The skill relies on executing shell commands to interact with version control and forge platforms. While it uses well-known tools (git, gh, glab), it also instructs the agent to run the prose through a tool named no-tropes, which is not a standard system utility and whose origin is unspecified in the skill.

pr-update