dev
Skill76 starsChanged 2 months agoMITView on GitHub
Load when implementing an issue or applying review fixes as a dev agent.
49 other projects use a version of this.
Adopted by
Most recently taken up by
- inancgumus/skillstheir own version, first seen Nov 8, 2026
- paulocsb/worldcup-fifa-pooltheir own version, first seen Nov 8, 2026
- fx/skillstheir own version, first seen Nov 8, 2026
- microsoft/Copilot-Managed-Runtimetheir own version, first seen Nov 8, 2026
- Mereithhh/very-happytheir own version, first seen Nov 6, 2026
- newkirkjosh/skillstheir own version, first seen Nov 1, 2026
Used with
Projects with it also use metrics-graphana, control-flow and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
slopusSkillmetrics-graphana
Query and manage Grafana dashboards and Prometheus metrics for Happy infrastructure. Covers grafanactl CLI usage, direct Prometheus queries through Grafana proxy, and dashboard-as-code workflows. Use when user asks about metrics, dashboards, monitoring, Grafana, Prometheus, or wants to add/modify panels.
Also in 3 projects that use this
slopusSkillcontrol-flow
Analyze and design control flows and data structures. Produces compact ASCII tree diagrams showing triggers, call chains, payload shapes, state mutations, and re-render effects. Use when user asks to diagram, trace, visualize, or design a flow or data structure.
Also in 3 projects that use this
MicrosoftSkillsessions
Core principles and workflow router for changes to the Agents Window under src/vs/sessions.
Also in 3 projects that use this
AnthropicSkilloffice-hours
Set up and run Claude Code office hours, including the calendar event when the champion's own calendar connector allows it. Use when a champion wants a regular time for teammates to bring problems.
Also in 3 projects that use this
nicepkgSkillteam
根据任务快速组建临时 AI Agent 团队协作。自动从 .claude/agents/ 中选择最合适的成员组队。
Also in 3 projects that use this
Microsoft AzureSkilltest
Also in 5 projects that use this
Other versions49 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 49
- slopus/happy★ 25ksame name
- microsoft/playwright-cli★ 15ksame name
- atopile/atopile★ 4ksame name
- cexll/myclaude★ 2.8ksame name
- waynesutton/markdown-site★ 628same name
- Automattic/wordpress-activitypub★ 583same name
- justrach/merjs★ 357same name
- stella/stella★ 274same name
- serithemage/serverless-openclaw★ 196same name
- Altman-conquer/agentrejoin★ 75same name
39 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository38 other files it gives its agents.
- AGENTS.mdAGENTS.md
- crates/AGENTS.mdAGENTS.md
- .github/AGENTS.mdAGENTS.md
- hooks/AGENTS.mdAGENTS.md
- pi-extensions/AGENTS.mdAGENTS.md
- skills/AGENTS.mdAGENTS.md
- tools/AGENTS.mdAGENTS.md
- ui/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- crates/CLAUDE.mdCLAUDE.md
- .github/CLAUDE.mdCLAUDE.md
- hooks/CLAUDE.mdCLAUDE.md
- pi-extensions/CLAUDE.mdCLAUDE.md
- skills/CLAUDE.mdCLAUDE.md
- tools/CLAUDE.mdCLAUDE.md
- ui/CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- app-deploySkill
- bot-instructionsSkill
- code-qualitySkill
- commit-guardsSkill
- deciderSkill
- deep-researchSkill
- dep-radarSkill
- doc-limitsSkill
- docs-writingSkill
- githubSkill
- iced-rsSkill
- kendex-issuesSkill
- linearSkill
- orchSkill
- preflightSkill
- price-handlingSkill
- project-managementSkill
- reviewerSkill
- review-gateSkill
- second-opinionSkill
- worktreeSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 31 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 47
Test shape (one control per surface, tables for shaped input, one file per surface) is code-quality § Tests.
- Distinctive: line 49
When a change deletes a call, apply code-quality § Cleanup to its callee.
- Distinctive: line 52
Before adding a function, parser, stub or loop, grep the repo for the verb it performs.
- Distinctive: line 54
Call or cite the one that exists, or escalate in your return.
- Distinctive: line 86
Pi.: Run it in the foreground.
“Common” is measured: kept by at least 6 in 10 of the 31 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- kendex:project-instructions:start -->Project Instructions
<!-- kendex:shared-instructions:start -->Problems with a kendex-owned skill go through kendex report; check ownership in the file first.
<!-- kendex:shared-instructions:end --><!-- kendex:project-instructions:end -->Dev Workflows
orch is the caller and runtime: it owns delegation format, round acceptance, and every shell-shape rule.
| Workflow | Purpose |
workflows/dev-implement.md | Implementation: activate → plan → implement → validate → commit → QA labels → summary → artifact → return (§ 1-11) |
workflows/dev-fix.md | Review fixes: evaluate → apply or skip → validate → commit → artifact → return |
Review and QA-review belong to the reviewer skill: ../reviewer/workflows/review.md, ../reviewer/workflows/qa-review.md. Command shapes are orch's ../orch/SKILL.md § Harness-Safe Shell; literal format tags and round mechanics are its ../orch/references/skill-rules.md § Format Tags Are Literal and § Round Closure.
Engineering Rules
- Scope is the issue's Done-when. A behavioral surface that does not trace to it stays out of this change, and a committed render of a source file you changed traces to whatever its source traces to. Two exceptions:
- the mechanical enablers of landing it ride without tracing to it: locks, changelog, baselines, dismissal renewals, that list and nothing else, never code that runs at runtime;
- a defect the change introduces or arms is in scope by definition, unless Step 0 of
../orch/references/finding-disposition.mdexcludes it.
- Every behavior change ships with a test that runs against the script or program enforcing it, at the smallest surface that fails. A workflow sentence ships no test. A test that pins prose, drives a second implementation, or stubs the function under test does not count as a test.
- A review finding adds a case only when it names a behaviour no existing case reaches. Otherwise it tightens the existing case's assertion, and the item reasoning names that case.
- A second fix round on the same function's guard is recurrence: redesign the rule under test so the class is unrepresentable, and fold the family of cases into one table.
- A test whose premise died is deleted whole in the commit that kills the premise, and the PR body names the deletion.
- Test shape (one control per surface, tables for shaped input, one file per surface) is code-quality § Tests.
- A refusal, a validator, a lock, a retry, or a test exists only for an input a real producer emits, this project's code or anything it calls or serves; name that producer beside it, or do not write it.
- When a change deletes a call, apply code-quality § Cleanup to its callee. Its deletion maps to the call removal's Done-when item; no internal caller is not proof that a supported external API is unused.
- A field, setting, or view member added by the change has a real producer and consumer. A named and documented external producer or consumer is valid when the change adds its in-repository counterpart; otherwise, add both sides in the change.
- No migration or compat code for this project's own formats, its manifest, settings, lock and cache shapes, never another tool's on-disk state, which an adapter may have to keep recognising: write no reader for an artifact an older version of this project wrote, and decline a finding that asks you to carry one forward. A layout, schema or cache change is one changelog line and a fresh install.
- Before adding a function, parser, stub or loop, grep the repo for the verb it performs; before stating a rule, grep for the rule.
- A second copy of that verb, in any language, is a twin and never delegation, and so is a second statement of a rule another file owns, in prose, config or a table.
- Call or cite the one that exists, or escalate in your return. An issue that orders a twin is escalated, not implemented.
- Docs move with the code they describe; the
docs-writingskill states the rule and thedoc-drift-checkhook names the docs that may need an update. - Once a pushed head has been reviewed, later rounds add commits and never amend; before any review has run on a head, the kendex-issues fix cycle may amend only to refresh a required check that cannot be rerun.
- A push that prints
rebase-map:lines has rewritten the shas the PR'sFixed in <sha>replies name: before holding, re-reply each such thread with the new sha, or post the map as one PR comment naming old and new per line; a sha the map reports asdroppedgets a reply that the fix commit no longer exists on the branch.
Code standards are ../code-quality/SKILL.md: correctness, comments, over-engineering, cleanup.
Round Contract
Execute workflow sections in order; a "Skip if" condition is the workflow's decision, never your own scope assessment. Never push and never open a PR. The orchestrator does that after review passes. A finding on a mechanism this diff introduces or arms is a fix whatever the round, unless Step 0 of the disposition flow excludes it; a Declined: there takes one of the reason forms ../orch/references/finding-disposition.md § Decision flow sets out, never a label or a test count.
The completion artifact is the round. dev-return-write writes it after the commit; never hand-author the JSON (schema: orch schemas/dev-return.md).
--issueis the delegation'sArtifact Key:line, the normalized workflow-state key (issue-Nfor GitHub,PROJ-123for Linear), never the tracker-nativeOWNER/REPO#Nor a bare number.--round-idis itsRound ID:line.--kindalways matches what was delegated.--validatematches your commit message and return.--validate-notecarries the test-only validation-ceiling report when that route applies. Flag constraints and value shapes:dev-return-write --help.
Acceptance is that artifact plus git state, never your message. Write the artifact, then return exactly once over the harness's agent-to-agent channel; a disk write is not a return. Send the **Return exactly** body once and go idle.
- The channel is Claude Code
SendMessage, Codexsend_input, OpenCode a resume on the storedtask_id, Pi background the final assistant message. - In a Pi persistent pane, follow the return with
complete_subagent; background agents must not call it. - On Codex the
send_inputMESSAGE is the durable return, and the runtime'sFINAL_ANSWERecho of it is expected, not a separate return to author or expand.
Validation
The validation gate and role ownership are complete in dev-implement.md § 5. Validate. Run no proof, rerun, receipt, isolation step, or approval step that section does not name. That section also owns the one proposed-rule route and the per-rule control for production gate and guard changes.
Long-Running Validation
Invariant, every harness: the completion tail (commit → QA labels → summary → artifact → return) is never dropped, and an interrupted run is never success. Re-check its real outcome and resume the tail. How you wait is your harness's:
- Claude Code. Background the BARE command with output redirected to a log via
run_in_background, never piped or chained, then wait for it with one bounded foreground poll: a for-loop oversleep 180with a cap, reading the[exited with code N]line that closes the task's own output file. Never idle for the completion notice and never depend on a background poller for it: the harness can kill your background shell on a low-memory heuristic that fires with free memory to spare, and the notice then never comes. The verdict is that exit code; the log holds command output and never an exit status. Then resume the tail. - Codex. Foreground and block.
- Pi. Run it in the foreground.
Reflect
Skip if nothing recurred and nothing surprised you. Otherwise put the lesson where it will be read again: architecture docs when an invariant, boundary or decision changed (the docs-writing skill says what belongs there), or the managing project's kendex config (kendex.toml at the kendex project root, kendex-local.toml in a source-catalog checkout) under [skill-instructions], [agent-additional-instructions], or [agent-launch-instructions]. Bar: would this save 5+ minutes in a future session? One surgical addition per lesson, no verbose examples. A config edit takes effect only once it is rendered, which you cannot do from a worktree, so name it, and anything else you cannot update yourself, in your return as [process] discovered work.
Configuration
Agent-type placeholders are project-configurable: [AGENT_TYPE] (dev agents receiving implementation delegations), [REVIEW_AGENT], [QA_AGENT]. Commit format: [PREFIX]([ISSUE_ID]): [DESCRIPTION]. DEV_VALIDATE_CMD (kendex.settings.toml [env]) names the project's full validation command for the Validate step; an empty value is the validation failure dev-implement.md § 5. Validate states, never a fallback.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

