The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a variety of system commands and local binaries to perform its tasks. These include git for version control snapshots, obsidian-wiki for linting and trust management, qmd for search indexing, and standard Unix utilities like rg, grep, sort, comm, and wc for data processing.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and audits markdown files from a local vault, which are untrusted data sources. These files could potentially contain malicious instructions or specially crafted content designed to manipulate the agent's behavior during the linting or fixing process.
Ingestion points: Markdown pages (.md), vault configuration files (AGENTS.md), and environment variables/config files (.env).
Boundary markers: None specified; the agent processes the raw content of the pages.
Capability inventory: The skill can write to the file system, execute powerful CLI tools (Git, QMD), and modify vault metadata.
Sanitization: There are no documented sanitization steps for the content read from the wiki pages before it is processed or used in logic.
[DYNAMIC_EXECUTION]: The skill provides explicit bash scripts for the agent to run, specifically for creating Git snapshots of the vault and refreshing the QMD search index. This involves multi-line script execution and the use of shell features like process substitution and environment variable expansion.

wiki-lint