The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted data from the project codebase (source code, configuration files, and documentation) to infer architectural decisions. It lacks boundary markers or sanitization to prevent potential instructions embedded in the codebase (e.g., in README.md or code comments) from influencing the agent's behavior.
Ingestion points: scripts/bash/setup-architect.sh and scripts/powershell/setup-architect.ps1 read content from across the entire project structure including package.json, requirements.txt, docker-compose.yml, README.md, and source files.
Boundary markers: Absent; the skill does not wrap ingested content in delimiters or include 'ignore' warnings for the model.
Capability inventory: The skill has the ability to write new ADR files to the .adlc/drafts/adr/ directory and execute project-specific setup scripts.
Sanitization: Absent; the skill relies on raw pattern matching and model inference without filtering the ingested text content.

architect-init