supervised-subagent
Use when the user asks to delegate implementation to a subagent
Skill by tenecsson
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/tenecsson-supervised-subagent
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/supervised-subagent
What's inside
- Skill name
supervised-subagent- Description
- Use when the user asks to delegate implementation to a subagent
- License
- MIT
- Instructions
- 22 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
Use one child for implementation and revisions while the parent owns acceptance. Workflow: 1. Spawn the child with a bounded task, acceptance criteria, relevant constraints, and proportionate verification. Honor the user's model and reasoning-effort choices; otherwise use the environment's defaults. 2. Tell the child to message the parent only when blocked or when an implementation or revision pass is complete. Do not request routine progress reports. 3. The parent may perform useful independent work. When none remains, call `wait_agent` with `timeout_ms: 1200000`, capped by tool limits and higher-priority instructions. Treat the wait as event-driven: it returns early for a child message or new user input. Do not shorten it merely to provide periodic updates. 4. After each completed pass, inspect the actual changes and run proportionate verification. Do not accept the child's summary as sufficient evidence. - Do not automatically require new tests or strict test-first sequencing. - Every new test should protect a plausible regression or durable contract. Reject tests that merely mirror constants, match source text, freeze private structure, or construct impossible internal states unless that is explicitly the required contract. - Honor user instructions to skip tests, worktrees, commits, documentation, or other process. 5. If changes are needed, send one consolidated, actionable revision request to the same child, then use the same wait. Repeat until material requirements and risks are addressed or a genuine blocker requires user input. Do not add rounds for speculative polish or assertion count. Polling rules: - Do not replace `wait_agent` with shell sleeps or another polling loop. - Before a timeout, do not call `list_agents` or send progress nudges. - After the first timeout, call `list_agents` once. If the child is active, give at most one concise user update and wait another interval without nudging. - Send a nudge only after a second consecutive timeout or evidence of inconsistent or stalled state. Report the accepted result, material review corrections, and verification evidence. If blocked, report the unresolved requirement and the input needed to continue.
From tenecsson/playbooks, MIT.
More from tenecsson/playbooks
- 0.529score
human-editor
tenecsson
Use when editing or reviewing existing prose for LLM writing habits, rhetorical padding, or fidelity to an author's voice.
SkillBFreshness: BClaude Code +9 - 0.524score
using-git-worktrees
tenecsson
Use when the user requests a worktree or parallel, experimental, or high-risk work needs isolation from the current checkout
SkillBFreshness: BClaude Code +9 - 0.524score
writing-skills
tenecsson
Use when creating, editing, or validating a reusable skill and you need to keep it concise, discoverable, and behavior-changing
SkillBFreshness: BClaude Code +9 - 0.523score
implementation-quality
tenecsson
Use when implementing or reviewing code changes that need judgment about proportionate testing and verification
SkillBFreshness: BClaude Code +9 - 0.519score
code-review
tenecsson
Use when the user requests code review, review feedback must be evaluated, or a high-risk change warrants an independent technical pass
SkillBFreshness: BClaude Code +9 - 0.518score
brainstorming
tenecsson
Use when a request has material product, behavior, or architecture choices that must be resolved before implementation
SkillBFreshness: BClaude Code +9 - 0.518score
planning-and-execution
tenecsson
Use when a multi-step implementation needs a durable plan or an existing approved plan needs coordinated execution
SkillBFreshness: BClaude Code +9 - 0.518score
systematic-debugging
tenecsson
Use when any bug, failing test, build break, or unexpected behavior needs root-cause analysis before fixes
SkillBFreshness: BClaude Code +9 - 0.493score
improve-coverage
tenecsson
Use when the user asks to improve test coverage or uncovered behavior needs a focused coverage pass
SkillBFreshness: BClaude Code +9 - 0.408score
dispatching-parallel-agents
tenecsson
Use when 2 or more tasks are independent enough to investigate or implement in parallel
SkillDFreshness: DClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 7, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/tenecsson/playbooks
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from tenecsson
- 0.529score
human-editor
tenecsson
Use when editing or reviewing existing prose for LLM writing habits, rhetorical padding, or fidelity to an author's voice.
SkillBFreshness: BClaude Code +9 - 0.524score
using-git-worktrees
tenecsson
Use when the user requests a worktree or parallel, experimental, or high-risk work needs isolation from the current checkout
SkillBFreshness: BClaude Code +9 - 0.524score
writing-skills
tenecsson
Use when creating, editing, or validating a reusable skill and you need to keep it concise, discoverable, and behavior-changing
SkillBFreshness: BClaude Code +9 - 0.523score
implementation-quality
tenecsson
Use when implementing or reviewing code changes that need judgment about proportionate testing and verification
SkillBFreshness: BClaude Code +9 - 0.519score
code-review
tenecsson
Use when the user requests code review, review feedback must be evaluated, or a high-risk change warrants an independent technical pass
SkillBFreshness: BClaude Code +9 - 0.518score
brainstorming
tenecsson
Use when a request has material product, behavior, or architecture choices that must be resolved before implementation
SkillBFreshness: BClaude Code +9
Cite this page
supervised-subagent skill. Agent Mart. https://agentmart.dev/skills/tenecsson-supervised-subagent (last updated 2026-10-24)