The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes git log to analyze file churn and historical changes, which it uses as a heuristic to identify the most critical and risky parts of the codebase during the intake phase.
[EXTERNAL_DOWNLOADS]: The skill provides instructions to the agent to offer the installation of additional modular skills from the author's collection (wondelai/skills/*) via the npx skills add command. These recommendations target vendor-owned resources and are presented as choices for the user rather than automated executions.
[PROMPT_INJECTION]: The skill ingests data from local project files and command-line outputs to manage its internal state and transitions between phases, creating a surface for indirect prompt injection.
Ingestion points: Reads state and history from docs/REMOVE-TECHNICAL-DEBT-PLAN.md, docs/TESTING.md, docs/TECH-DEBT.md, docs/ARCHITECTURE.md, and docs/RELIABILITY.md, as well as output from git log.
Boundary markers: The instructions do not define delimiters or provide specific directives to ignore embedded instructions within these ingested files.
Capability inventory: The skill has the capability to write to the local filesystem, invoke other platform skills, and generate shell commands for user execution.
Sanitization: No validation or sanitization is performed on the content retrieved from the project's documentation before it is used to influence the agent's workflow.

remove-technical-debt