Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add GongShichen/LuminaCode --skill task-breakdowngit clone --depth 1 https://github.com/GongShichen/LuminaCodeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/gongshichen/luminacode/task-breakdown)<a href="https://agentmods.dev/skills/gongshichen/luminacode/task-breakdown"><img src="https://agentmods.dev/badge/skills/gongshichen/luminacode/task-breakdown/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/gongshichen/luminacode/task-breakdown"><img src="https://agentmods.dev/badge/skills/gongshichen/luminacode/task-breakdown.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00020 | $0.00881 |
| Opus 5.5 | $0.00008 | $0.00352 |
| Sonnet 5.5 | $0.00004 | $0.00176 |
| Haiku 4.5 | $0.00002 | $0.00088 |
Grade A, and why
Task Breakdown scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Create a task graph that can actually drive the Team Loop.
Inputs:
- User request and current working directory.
- Existing repository/project structure.
- Required product-development stages from the shared prompt.
- Current dialogue, active A2A tasks, gate status, and known artifacts.
Task graph structure:
-
Objective
- One-sentence user outcome.
- Completion criteria in observable terms.
- Non-goals and explicitly deferred scope.
-
Project/artifact root
- If the user asks for a new named project or directory and gives no absolute path, set the root to
<current working directory>/<requested name>. - All owner tasks, file paths, README paths, tests, and gate checks must use that root.
- Add a parent-workspace-clean check for named projects.
- For Python projects, decide both the product/display name and the importable package/module name. Hyphenated names such as
mini-tasksare product names, not validpython -mmodule names. Unless packaging creates a console script, contract commands must use the importable name such aspython -m mini_tasks. - For CLI/TUI/local tools with persistent data, decide how runtime data paths are resolved and how subprocess tests isolate storage. Pick one concrete mechanism visible to real processes, such as an environment variable, cwd convention, config file, fixture path, or command flag. Do not list monkeypatch as a subprocess integration mechanism.
- If the user asks for a new named project or directory and gives no absolute path, set the root to
-
Required stage artifacts
- PRD:
PRD.md. - UX design:
UX_DESIGN.md. - Backend plan:
BACKEND_PLAN.md. - Frontend plan:
FRONTEND_PLAN.md. - Shared interface contract:
INTERFACE_CONTRACT.md. - Integration evidence:
INTEGRATION_REPORT.md. - QA report:
QA_REPORT.md. - Reviewer report:
REVIEW_REPORT.md.
- PRD:
-
Components and ownership
- Frontend/user-facing surface owner.
- Backend/platform/data/integration owner.
- For CLI/TUI/local-tool products, assign command parsing, output formatting, help/error copy, and interaction behavior to Frontend; assign data model, business services, persistence, package manifests, storage paths, and backend tests to Backend.
- UX owner.
- QA and Reviewer gates.
- DevOps or Research support only when needed.
-
Dependency order
- PRD before UX.
- UX before frontend/backend technical plans.
- Technical plans and interface contract before implementation.
- For persistent local tools, the interface contract must include storage path resolution and subprocess-visible test isolation before implementation.
- Implementation before integration.
- Integration before QA/Reviewer pass.
- QA/Reviewer findings before final completion.
-
Dispatch packets
- Each specialist task must include project root, input document paths, expected output document path, constraints, and acceptance evidence.
- Avoid assigning one task that crosses role boundaries; split it.
- If a target already has a running/pending A2A task, do not duplicate the task.
- Required commands must be runnable as written and preserve exit codes. Do not put
| head,| tail, or other output truncation in acceptance commands unless the status is explicitly preserved.
-
Recovery path
- Missing artifact: dispatch the owner to write it.
- Wrong root: dispatch cleanup/move work.
- Contract mismatch: dispatch Frontend and Backend to reconcile
INTERFACE_CONTRACT.md. - Implementation/test failure: identify the owning boundary, then dispatch a narrow repair task to that specialist. Team Leader should not edit product source, tests, package manifests, runtime data, or generated build artifacts directly.
- QA fail or Reviewer reject: dispatch concrete repair tasks to the owner, then regate.
Clarification policy:
- Ask the user only when a decision cannot be inferred safely from the request, repository, or reasonable product defaults.
- Record assumptions in PRD or Team Acceptance Contract when proceeding.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 76 lines · 20 tokens per session scan A 7e9b27804bc4
Task Breakdown is a skill published in the GitHub repository GongShichen/LuminaCode (9 stars, last pushed 4d ago), licensed Apache-2.0. It adds 20 tokens to every session and 881 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
story-readiness
Checks whether a story is ready to implement: clear acceptance criteria, no open questions, valid ADR and dependencies. Use when the user asks if a story, or the sprint's stories, are ready to start or build.
recipe-create-meet-space
Create a Google Meet meeting space and share the join link.
workthreads
SpecStory Workthreads - a weekly work-thread rollup across a team's repos from SpecStory coding histories (any agent - Claude Code, Codex, Cursor, Gemini, and more). It groups the window's sessions into threads of work per project and labels each new / open / recently closed, so a lead sees what shipped, what is still…
atmos-config
Atmos root configuration: atmos.yaml discovery, precedence, deep merging, basepath, imports, minimal bootstrap, and routing to narrower Atmos skills.
qa-project-bootstrap
Onboard a new QA engineer to an existing codebase, or audit an existing test architecture. Produces a 30-day ramp plan: codebase orientation, framework walkthrough, test architecture audit, mentorship pairing, and first-test guidance. Use when: "QA onboarding," "new tester," "ramp up," "test architecture audit,"…
remove
Remove a deployed framework or addon from the current workspace.