cli-sub-agent

Use when: full dev cycle branch->plan->implement->review->PR->merge

Use when: executing tasks via csa run/review/debate, session mgmt

Use when: iterative review-fix loop until csa review --diff is clean

Use when: running CSA-driven code review, independent model selection

Use when: three-layer manager-employee orchestration for delegation

Use when: legacy alias, redirects to dev2merge pipeline

Use when: converting TODO plan or open GitHub issues into deterministic execution checklist

Use when: creating TODO plan with debate-enhanced CSA reconnaissance, including epic decomposition for large tasks

Use when: adversarial multi-tool debate for strategy, tier escalation

Split monolith project documentation (CLAUDE.md, etc.) into condensed summary + reference detail files following AGENTS.md pattern

Use when running a non-blocking CSA background code health scan that uses csa health and csa tokuin estimate to propose refactoring GitHub issues for files over token or complexity thresholds.

Use when: PR review loop with cloud bot, arbitration, and merge

Use when: committing code with security audit, tests, review gates

Recover main-agent context after `/clear`, `/compact`, or lost local thread state by using `csa recall` against recorded Claude main sessions.

Use when: merged PR had HIGH/CRITICAL findings that represent a bug class — extracts reusable coding rule

Use when: review found 2+ independent findings in different files, fix phase can parallelize RECON

Use when: committing with pre-commit review, authorship-aware fix-retry

Launch long-running commands (>4 min) via nohup and poll at kv_cache.long_poll_seconds (default 240s) to keep KV cache warm.

Use when: deep per-crate code audit, Chinese documentation generation

Use when: creating, auditing, or optimizing repository quality gates (git hooks, pre-commit checks, CI guards, merge protection)

Use when: reviewing GitHub PR via gh CLI, scale-adaptive strategy

Use when: creating CSA pattern (PATTERN.md + skill + workflow.toml)

Use when: generating technical blogs from codebase-audit results

Use when: CSA hits operational error, filing structured GitHub issue

Use when: per-file AGENTS.md compliance audit with Chinese reports

Use when: pre-commit security audit (tests, vulns, code quality)

Reviewer CSA for fact-checking Writer outputs against source code

Writer CSA for deep crate analysis — generates README, review report, blog, and facts.toml

Run CSA project migrations to update config and workflow files to the current version

Expert diagnosis for Tokio/async Rust issues including deadlocks, task leaks, performance bottlenecks, and cancellation safety

Clear, practical documentation for APIs, README, architecture decisions, and changelogs

Comprehensive guide for professional Rust development combining implementation expertise, architectural depth, and project standards

Adversarial security analysis expertise for identifying vulnerabilities before attackers do

Comprehensive test design for Rust projects following Core-first TDD and test pyramid principles