The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions in SKILL.md direct the agent to execute a local Python script scripts/validate_brief.py to validate a JSON handoff file. This is an intentional functional design for ensuring data integrity and handoff quality.
[EXTERNAL_DOWNLOADS]: The skill documentation references an official vendor repository (github.com/srinitude/skills) to check for prerequisite skills like starting-point. As these resources belong to the skill's official vendor and are documented, they are considered safe.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an inherent attack surface for indirect prompt injection due to its core function of ingesting and processing user-supplied thoughts and resuming state from project-local markdown log files. 1. Ingestion points: User input via the new command and state restoration from reify-log-*.md files. 2. Boundary markers: The skill employs a structured markdown record format for state preservation but lacks explicit delimiters or instructions to ignore potential commands embedded in the processed data. 3. Capability inventory: The agent has the capability to execute specific shell commands (validation scripts) and perform file system operations (copying templates, moving probe files). 4. Sanitization: No explicit sanitization or filtering of the interpolated user content is performed before it is written to logs or processed.
[DYNAMIC_EXECUTION]: The skill's test suite in scripts/tests/test_validate_brief.py utilizes the subprocess module to execute Python scripts. This behavior is confined to the skill's internal validation tests and follows standard development practices for testing command-line utilities.

reify