The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run /setup-matt-pocock-skills to configure conventions and triage vocabularies. This is a non-standard command invocation that executes logic outside the skill's own code.
[COMMAND_EXECUTION]: The skill uses shell-based CLI tools (gh for GitHub and glab for GitLab) to create issues. It uses heredoc patterns to pass issue bodies, which may be vulnerable to command injection if the source material (plans/PRDs) contains shell-breaking characters that are not properly escaped by the platform's tool-calling implementation.
[PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection (Category 8) because it processes untrusted external data (plans, specs, or PRDs) to generate instructions for 'AFK agents'.
Ingestion points: Reads plan, spec, or PRD content provided in the conversation context or fetched from external issue trackers (Step 2).
Boundary markers: None. The skill does not use delimiters or instructions to ignore embedded commands within the processed data.
Capability inventory: The agent can execute shell commands (gh, glab) and write files to the local file system (.scratch/).
Sanitization: There is no programmatic sanitization or escaping of the content. While Step 5 includes a user review 'quiz', this manual checkpoint relies on the user spotting malicious instructions or formatting that could lead to command injection or downstream agent manipulation.

to-issues