The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a local binary utility to perform repository inventory tasks.
Evidence: SKILL.md and scripts/platform_inventory.py instruct the agent to execute /home/bjorn/.codex/skill-support/bin/repo-inventory with various flags.
The utility is located in the author's home directory (/home/bjorn/), which corresponds to the skill author 'BjornMelin'.
[PROMPT_INJECTION]: The skill instructions create a surface for indirect prompt injection by prioritizing instructions found within the analyzed repository.
Ingestion points: The agent is instructed to read AGENTS.md from the target repository first for 'local constraints and canonical commands'.
Boundary markers: No explicit boundary markers or instructions to ignore embedded malicious prompts are provided when reading AGENTS.md.
Capability inventory: The skill has the capability to execute subprocesses (via platform_inventory.py) and access external documentation (via context7 and exa MCP tools).
Sanitization: No sanitization or validation of the content within AGENTS.md is performed before the agent adopts its instructions as constraints.

platform-architect