The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources such as issue bodies, design notes, and previous run reports to generate its decision list.
Ingestion points: Decisions are collected from session history, issue trackers, and project artifacts as specified in SKILL.md Step 1.
Boundary markers: The skill does not explicitly define markers to separate untrusted data from its internal instructions when forming dialogs.
Capability inventory: The skill uses Write and Edit tools to modify project files and Bash to process data.
Sanitization: Although it scans for secrets, it does not detail specific sanitization for potential instructions embedded in the ingested markdown or prose.
[EXTERNAL_DOWNLOADS]: The skill's documentation references a configuration schema at "https://raw.githubusercontent.com/TitusKirch/skills/main/tituskirch-skills.schema.json". This is a vendor-owned resource used for configuration structure definitions.
[COMMAND_EXECUTION]: The skill executes a local shell script, "templates/resolve-config.sh", and uses jq via the Bash tool to resolve and merge configuration profiles from the project's ".tituskirch-skills.json" file.

decide