The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from several external sources that could contain malicious instructions intended to influence the agent's behavior during the reflection or memory-update phases.
Ingestion points: The skill reads and aggregates Slack messages (references/agent-gather-slack.md), GitHub commit messages and pull request reviews (references/agent-gather-github.md), and Fireflies meeting transcripts (references/agent-gather-fireflies.md).
Boundary markers: The instructions do not define clear delimiters or include safety warnings to the sub-agents to ignore potential instructions embedded within the fetched data.
Capability inventory: The skill possesses the ability to write new files (reflection journals) to the filesystem and modify the agent's persistent long-term memory (SKILL.md Steps 4 and 5).
Sanitization: No evidence of sanitization, escaping, or filtering is present for the data gathered from external platforms before it is interpolated into the analysis prompts.
[COMMAND_EXECUTION]: The skill requires the user to execute a local Python script for configuration management.
Evidence: The setup instructions in SKILL.md prompt the user to run python <skill-dir>/scripts/skillctx-resolve.py resolve p-daily-reflection to resolve internal bindings and python <skill-dir>/scripts/skillctx-resolve.py set ... to update them.

p-daily-reflection