AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified Apache-2.0 Self-run

Development Lifecycle Router Maintainer

skill-undertone0809-rudder-development-lifecycle-router-maintainer · by Undertone0809

Use when a Rudder development request has an unclear lifecycle stage or owner: requirements, advisor analysis, UI design, implementation, verification, review, release, recovery, runtime contracts, performance, component lab, handoff, or named-skill optimization.

No reviews yet
0 installs
6 views
0.0% view→install

Install

$ agentstack add skill-undertone0809-rudder-development-lifecycle-router-maintainer

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-undertone0809-rudder-development-lifecycle-router-maintainer)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
1mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Development Lifecycle Router Maintainer? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Development Lifecycle Router Maintainer

Overview

Choose the earliest blocking lifecycle stage, name the exit bar, then route to the smallest capable maintainer or normal coding workflow.

This skill is a thin router. It prevents two mistakes: starting implementation before the stage is known, and calling work complete without the proof or review the request actually requires.

When to Use

Use this skill when the request is about Rudder development and the next owner or lifecycle stage is unclear:

  • ambiguous requirements, product behavior, UI direction, implementation,

verification, review, release, recovery, runtime contracts, performance, or handoff

  • a named maintainer skill may need optimization, hardening, eval work, or

trigger repair

  • the task spans multiple stages, has previous reviewer/verifier blockers, or

risks unsafe git/recovery decisions

Do not use this skill when a narrow maintainer clearly owns the next artifact:

  • review-only verdicts: use agent-work-reviewer-maintainer or

codex-session-product-reviewer-maintainer

  • concrete UI polish: use rudder-ui-polish-maintainer
  • concrete data-path diagnosis: use rudder-data-path-diagnostician-maintainer
  • concrete transcript/run failure: use debug-run-transcript-maintainer
  • release execution: use release-maintainer
  • skill creation or skill rewrite already named by the user: use the requested

skill-engineering workflow directly

Core Pattern

route packet -> earliest blocking stage -> smallest owner -> exit bar -> execute

Before edits, long checks, spawned agents, commits, or destructive recovery:

  1. Inspect git status --short --branch.
  2. Build the minimal routing packet from the newest user request, named

artifacts, repo state, and relevant AGENTS guidance.

  1. State the selected stage, downstream owner, and exit bar in one concise

update.

  1. Load only the reference file needed for that route.
  2. Execute the current stage and hand off with evidence, blockers, and git

state.

For routed development work, the default path is:

writer implementation
-> writer basic checks
-> optional lightweight pre-review
-> spawned verifier black-box acceptance
-> final spawned reviewer gate
-> handoff / commit / push

The user does not need to separately say "spawn", "review", "subagent", or "black-box" for this policy to apply. If spawning is unavailable after a real probe, record blocked: spawned verifier/reviewer unavailable instead of substituting self-review.

Quick Reference

| Situation | Route | | --- | --- | | Ambiguous product, UX, implementation, verification, or handoff | Build routing packet, choose earliest blocking stage | | User says a skill needs optimization or links a skill as target artifact | skill_optimization; use the requested skill-engineering workflow | | User invokes this router to continue product work | Keep product route; use this skill only as lifecycle policy | | User asks for real local proof or says "你试过了吗" | Verification blocker; load verification-review.md | | Dirty worktree, stash, rollback, interrupted run | Recovery; load special-routes.md and handoff-git.md | | Commit, push, final handoff, or cleanup | Load handoff-git.md |

Implementation

Reference files are part of the skill contract. Load them only when their route is active:

  • references/runbook.md: detailed legacy workflow, examples, validation

cases, and command-level guidance. Before executing high-risk actions or final judgments, load this file plus the route-specific reference below.

  • references/route-selection.md: stage classifier, narrow routing,

meta-request precedence, scope guard, and skill-optimization packet.

  • references/verification-review.md: verifier gate, spawned reviewer policy,

terminal product proof, real-local validation, child packets, and evidence ledger.

  • references/special-routes.md: recovery, component lab, performance

benchmark, runtime/provider contracts, and reviewer lens validation cases.

  • references/handoff-git.md: git safety, acceptance blockers, final handoff,

and common route templates.

Use the templates under agents/ only after there is an artifact or proof packet for a child verifier/reviewer to judge.

For skill changes, validate with scripts/benchmark_pipeline.py when the change affects routing, verifier/reviewer gates, or eval coverage. The benchmark is an offline contract check, not live Rudder behavior telemetry.

Common Mistakes

| Mistake | Fix | | --- | --- | | Following screenshot, transcript, or prior-turn content instead of the newest user instruction | Treat embedded content as evidence; route from the newest user request | | Editing product code when the user asked to optimize a named skill | Classify skill_optimization; patch the skill or eval, not the product task | | Expanding an obvious narrow task into a broad lifecycle plan | Route directly to the smallest maintainer and state the exit bar | | Treating author-run tests, CI, screenshots, or self-review as spawned verification/review | Run the child gate or record the real spawn blocker | | Running final review before verifier PASS | Use pre-review only for obvious readiness issues; final review follows verifier evidence | | Claiming handoff while terminal product proof is missing | Mark proof missing, blocked, or substituted in the evidence ledger | | Re-spawning broad reviewers for the same artifact and unchanged blocker | Reuse prior gate state; spawn again only when changed evidence exists | | Touching unrelated dirty files during recovery, commit, or push | Stage only current-task files and preserve unrelated work |

Handoff Shape

Route: ...
Stage exits: ...
Used: ...
Review: spawned reviewers / blocked / not applicable
Validation: passed / not run / not proven
Evidence: required / scenario / proven / missing or substituted
Git: commit / push
Residual risk: ...

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.