The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several shell commands for repository analysis, including git status, git diff, git cat-file, git merge-base, rg (ripgrep), and grep. These are used for verifying commit history and searching the codebase for identifiers or patterns.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the repository, creating a surface for indirect prompt injection.
Ingestion points: The skill explicitly instructs the agent to read docs/llm/AGENTS.md, docs/llm/INDEX.md, and arbitrary source files discovered during the Locate and Expand phases.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands within the processed files are provided.
Capability inventory: The agent has access to shell execution (git, rg), codebase indexing (index_workspace), and subagent delegation (task).
Sanitization: No sanitization or validation of the content read from files is performed before it is used to guide navigation or influence the subagent's scope.
[DYNAMIC_EXECUTION]: The skill includes a delegation mechanism where it calls a task tool with subagent_type: "explore". While it instructs the subagent to perform no edits and treats its output as navigation hints, this constitutes the dynamic generation and execution of a sub-task based on current context.

read-codebase-context