The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data during its research and audit phases, creating a surface for indirect prompt injection.
Ingestion points: The skill reads the entire scope of files provided for research or auditing as defined in Step 1 of SKILL.md.
Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" markers when reading untrusted file content.
Capability inventory: The skill utilizes shell commands via git (in references/code-review.md) and network fetching via curl or similar tools (in references/source-ladder.md).
Sanitization: No sanitization, filtering, or validation of the processed content is specified before it is incorporated into the agent's context.
[REMOTE_CODE_EXECUTION]: The skill fetches supplementary instruction sets from a remote repository.
Finding: The agent is instructed in references/source-ladder.md to fetch a specific skill definition from https://raw.githubusercontent.com/D1ZZY4/agent-skills/refs/heads/main/skills/context7-expert/SKILL.md if it is not already installed.
Context: The resource is located in the official repository of the skill's author (D1ZZY4). This represents the dynamic loading of instructions from a vendor-controlled remote source.
[COMMAND_EXECUTION]: The skill executes shell commands to perform technical code reviews.
Finding: references/code-review.md uses git rev-parse, git diff, and git log to analyze changes between a fixed point and the current repository state.
Context: These are standard operations for code review tasks and are restricted to the local repository context.
[EXTERNAL_DOWNLOADS]: The skill is designed to fetch information from a variety of external sources to verify claims.
Finding: references/source-ladder.md defines a hierarchy for fetching data from live official documentation, repositories, and registries.
Safety: The skill includes explicit safety boundaries in SKILL.md to avoid transmitting secrets or proprietary code during these network operations.

deep-research-expert