Agent Claude Code
Surgical 1-2 file edit. Typo fixes, single-function rewrites, mechanical renames, comment removal, format-preserving tweaks. Hard refuses 3+ file scope. Returns caveman diff receipt. Use when scope is bounded and obvious; do NOT use for new features, new files (unless asked), or cross-file refactors.
Agent Claude Code
Read-only code locator. Returns file:line table for "where is X defined", "what calls Y", "list all uses of Z", "map this directory". Output is caveman-compressed so the main thread eats 60% fewer tokens than vanilla Explore. Refuses to suggest fixes.
Agent Claude Code
Diff/branch/file reviewer. One line per finding, severity-tagged, no praise, no scope creep. Output format path:line: : . . Use for "review this PR", "review my diff", "audit this file". Skips formatting nits unless they change meaning.
Agent Claude Code
Independent Argus PR reviewer. Executes the canonical /review contract (staged gates, evidence ledger, report + ledger receipt) for a pull request. Uses opencode-go/deepseek-v4-flash for review.
Agent Claude Code
LUNA-model execution subagent (openai/gpt-5.6-luna). Use for /fix, /pr, implementation, and execution of an approved plan in Argus harness workflows.
Agent Claude Code
SOL-model reasoning subagent (openai/gpt-5.6-sol). Use for orchestrator, planning, review, design, and any thinking-heavy analysis in Argus harness workflows.
Agent Claude Code
Wave build agent for any wave loop. Implements the assigned task in an isolated worktree, commits, and pushes the branch. Used for parallel task execution across waves. Uses openai/gpt-5.4 for build work.
Agent Claude Code
Wave review agent for any wave loop. Reviews diffs line-by-line and posts inline GitLab MR comments with glab on the exact changed lines. Used in the build-review-fix loop after task subagents finish. Uses openai/gpt-5.4 for review.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: