bahayonghang/claude-code-companion
content/skills/developer-tools-integrations/claude-code-companion/SKILL.md
Coordinate Claude Code companion-style workflows for multi-step implementation, bounded review, follow-up execution, and session-to-session continuation inside Claude Code. Use this skill when the user explicitly wants a Claude Code-native companion workflow, wants help structuring background or follow-up work in Claude Code, or needs clear guidance on when to use direct Claude Code actions versus a persistent companion-style process. Prefer this over generic advice whenever the task is specifically about Claude Code orchestration, review-first execution, or continuing a previous Claude Code task safely.
$ install --global
skillsauthnpx skillsauth add bahayonghang/my-claude-code-settings claude-code-companionInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 21, 2026, 3:11 AM221.3s2 files scanned
SKILL.md
- name:
- claude-code-companion
- description:
- >-
- version:
- 0.1.0
- category:
- developer-tools-integrations
- argument-hint:
- [task-or-workflow-goal]
Use this skill when the user wants a Claude Code-native companion workflow rather than a one-off command.
Best fit
Reach for claude-code-companion when the user wants one or more of these outcomes:
- break a Claude Code task into review-first, implementation, and verification stages
- continue a prior Claude Code effort without overpromising hidden persistent threads
- coordinate bounded follow-up work after an earlier review or diagnosis
- decide when work should stay in the foreground versus be structured as a follow-up process
- standardize how Claude Code handles multi-step task continuation inside the current repository
Positioning
| Skill | Best for |
|---|---|
| claude-code-companion | Companion-style orchestration inside Claude Code itself |
| codex-companion | Codex-specific background jobs, resumable Codex task threads, and runtime lifecycle commands |
| direct Claude Code usage | Simple one-shot reads, edits, tests, and short implementation loops |
Working model
Claude Code does not automatically guarantee the same runtime model as Codex app-server threads. Treat this skill as a workflow contract, not a promise of Codex-style background infrastructure.
When using this skill:
- Clarify the user goal and success criteria.
- Decide whether the task needs:
- direct execution now,
- a review-first flow,
- a staged follow-up flow,
- or a resume-from-context flow.
- Prefer the smallest safe next step.
- Keep claims precise: if a task is being resumed from repository context rather than a provider-native persistent thread, say so explicitly.
- Before claiming completion, verify the affected code paths, docs, or tests.
Suggested flow
1. Review-first flow
Use this when the user wants validation before edits.
- Inspect the relevant diff, files, or behavior
- Summarize key risks
- Confirm whether implementation should proceed
- Apply changes only after the user wants action
2. Follow-up implementation flow
Use this when the user already approved a direction.
- restate the concrete target
- identify the minimum file set to touch
- implement in small steps
- run targeted verification
- report what changed and what still needs review
3. Continuation flow
Use this when the user says “continue”, “pick up where we left off”, or similar.
- recover the target from current repo state, task context, and prior conversation context
- explicitly say what is being resumed
- do not imply hidden provider-native thread continuity unless you actually have it
- pick the next highest-value bounded step
Guardrails
- Do not imply Codex-style persistent task threads unless the underlying Claude Code workflow actually provides them.
- Prefer explicit review and verification over ambiguous “background magic”.
- Keep scope tight; if a task expands, reframe it before continuing.
- If the user needs Codex-specific runtime lifecycle commands like
status / result / cancel, route them tocodex-companioninstead.
References
If the task needs deeper Claude Code usage guidance, also inspect nearby repository guidance and the project’s Claude-related docs before deciding on the workflow.
Related Skills
bahayonghang/goal-meta-skill
development
VerifiedTrustedCommunity
Turn vague or complex Codex tasks into strong `/goal` commands with outcome, verification, constraints, boundaries, iteration policy, completion evidence, and pause/block conditions. Use when the user asks for Codex goal instructions, Goal 指令, 目标指令, `/goal` prompts, 中文 Goal 模板, plan-to-goal interviews, success criteria, verification commands, or bounded agent work definitions.
15SKILL.mdUpdated Jun 13, 2026
bahayonghang/ast-grep
tools
VerifiedTrustedCommunity
Write, debug, and validate ast-grep structural code search rules. Use this skill when the user needs syntax-aware code search, AST pattern matching, structural refactor discovery, language-construct queries, or searches that plain text tools like rg can miss, such as finding functions with particular descendants, calls inside specific contexts, missing error handling, React hook shapes, decorators, or other Tree-sitter-backed code structures.
15SKILL.mdUpdated Jun 9, 2026
bahayonghang/goudi
development
VerifiedTrustedCommunity
Use when the user asks to ground an ambitious proposal, avoid over-grand designs, make a bold direction executable, pressure-test feasibility, prevent "too much vision and too little landing", or turn a strategy/refactor/product idea into the smallest verifiable first move with stop rules. Trigger for requests such as 落地, 先落地, 别太飘, 收一收, 可执行, 可验证, 止损, and for follow-ups after geju-style big-picture thinking. Do not trigger for ordinary code review or implementation unless the user explicitly asks to ground or shrink the plan first.
15SKILL.mdUpdated Jun 7, 2026
bahayonghang/geju
development
VerifiedTrustedCommunity
Use when the user explicitly asks to think bigger, open up the design space, challenge conservative design, avoid over-indexing on backward compatibility, escape local-detail fixation, or make a bold high-level product or architecture direction call. Use for strategic reframing, not for ordinary code review, PRD writing, implementation planning, or adversarial risk review.
15SKILL.mdUpdated Jun 7, 2026