Compound Engineering is a plugin of 36 skills for AI coding agents. It runs a loop of brainstorm, plan, work, simplify, review and compound, and writes each change's lessons where the next run can read them. It is for developers who use Claude Code, Cursor, Codex and other agent hosts.
Why read it
When a skill runs inside another workflow, it ends with a status drawn from a fixed list of words, and it is told never to rename them. The calling skill can then branch on a known value instead of reading free prose, so a chain of skills keeps working without a person in the loop.
Tokens like `mode:pipeline`, `mode:return-to-caller` and `mode:non-interactive` tell a skill that no person is there to answer. Each question the skill would normally ask is replaced by a cautious default, so the same skill works both by hand and inside an automated run.
Each phase names the reference file it needs. A read made before that phase does not count, and if a file cannot be read the skill stops instead of guessing its rules. This keeps the main file small and prevents the agent from working from a half-remembered version of the rules.
Most skills open with an Outcome line and a Done line that say exactly when the run is complete. Some also name the cases where the right result is to write nothing. This gives the agent a clear point to stop.
The rule for finding the artifact root is pasted into many skills between `ce-docs-root:start/end` comment markers. ce-setup holds the main copy. Each skill gets the full rule, and the markers show the copies are meant to stay the same.
ce-compound decides whether to record a learning with one question: if this note disappeared, would a future engineer still repeat the mistake? The question has a clear answer, so the skill does not record routine work, and an explicit request does not lower the bar.
Skills sort the request into tiers, such as a short chat answer, a brief or a full document. When the agent cannot tell which tier fits, it must take the heavier one. Small requests stay small, and risky ones are not under-handled.
ce-noslop turns its writing rules into numbered questions the agent asks of every sentence. One example: could this sentence move to another project unchanged? The same tests work as rules while writing and as checks while editing or detecting.
A model read each skill to find these. Top to bottom is the usual order. Select a box to read the skill, or a line to see the text behind it.
Links to many skills, so not drawn: ce-bakeoff (7), ce-brainstorm (10), ce-debug (9), ce-explain (7), ce-plan (9), ce-pov (10), ce-work (10), lfg (16)
Many skills link to it, so not drawn: ce-brainstorm (11), ce-code-review (6), ce-commit-push-pr (6), ce-compound (8), ce-debug (13), ce-noslop (11), ce-plan (8), ce-pov (8), ce-work (7)
ce-compound-refresh, ce-doc-review, ce-optimize, ce-promote, ce-prototype, ce-riffrec-feedback-analysis, ce-simplify-code, ce-sweep, ce-test-xcode, wtf
Every one is a skill.
| SKILL | STARTED BY | DESCRIPTION | LINKS |
|---|---|---|---|
| ce-babysit-pr | You or the agent | Babysits an open GitHub PR until merge-ready, first judging whether an old PR still belongs. Use when asked to watch a PR over time or update an old one — not for one-shot comment resolution or one CI failure. GitHub (incl. Enterprise) only. | 6 |
| ce-bakeoff | You or the agent | Develop independent competing solutions to a defined brief, compare them, and synthesize a winning approach. Use when choosing well requires developing alternatives beyond their current form. Use ce-pov to judge developed material and ce-ideate to discover opportunities. | 7 |
| ce-brainstorm | You or the agent | Explore vague or ambitious ideas into a right-sized requirements-only unified plan. Use when the user wants to brainstorm or scope what to build. Not for executing already-specified work. Use ce-pov for a verdict on adopting a named external technology. | 13 |
| ce-code-review | You or the agent | Review a named diff or PR for bugs, regressions, tests, and standards. Use when asked to review code or when a shipping skill needs a review receipt. Use when asked to apply this review's findings locally. Use ce-resolve-pr-feedback for feedback already left on a PR. | 6 |
| ce-commit | You or the agent | Create a git commit with a clear, value-communicating message. Use when the user asks to commit/save staged or unstaged changes with a repo-appropriate message. | 2 |
| ce-commit-push-pr | You or the agent | Commit, push, and open a PR. Use when asked to ship/open a PR, or for PR-description-only flows like writing, rewriting, or describing a PR body. | 5 |
| ce-compound | You or the agent | Document a solved problem as a durable repo learning. Use when verified work produced non-obvious reasoning absent from its final code, tests, or existing docs; avoid routine fixes whose artifacts already explain the lesson. | 2 |
| ce-compound-refresh | You or the agent | Refresh the repo's captured learnings against the current codebase. Use when auditing stale, overlapping, superseded, or drifted learnings; avoid general refactor, debugging, or code review unless the learnings store is explicit. | 1 |
| ce-debug | You or the agent | Diagnosis loop for bugs and failing behavior. Use when asked to debug or fix failing or slow behavior. | 10 |
| ce-doc-review | You or the agent | Review requirements, plans, or specs with role-specific lenses. Use when the user wants to improve an existing planning document. | 4 |
| ce-dogfood | Only you | Hands-off, diff-scoped browser QA of the active branch: maps user flows, drives a real browser, autonomously fixes small breakages with regression tests and commits, judges experience against product personas, and writes a durable dogfood report. Manual invocation only. | 5 |
| ce-explain | You or the agent | Explain how and why something has its current shape, or what happened over a window of work, grounded in evidence. Use when the user asks for an explanation. Use ce-pov for a judgment or recommendation. | 7 |
| ce-handoff | You or the agent | Create a session handoff for another agent, or resume, find, and read any user-selected continuity source. Use when work or conversation must continue without access to the current session history. | 1 |
| ce-ideate | You or the agent | Generate and evaluate grounded ideas. Use when the user wants ideas, improvements, or surprising directions before choosing one to develop. Not for refining an idea they already have (ce-brainstorm) or judging one already on the table (ce-pov). | 6 |
| ce-noslop | You or the agent | Rewrite, check, or draft prose so it carries no AI writing tells, reads plainly on the first read, and keeps every source fact. Use when asked to make writing plainer or free of those tells, to check writing for them, or when drafting from supplied content. Use ce-promote for channel-specific marketing copy. | 1 |
| ce-optimize | You or the agent | Optimize a named target with a measured loop: attribute a workload's cost, or score variants and keep winners. Use when a working system's metric should move and the winning change is not already known. Use ce-debug when the job is diagnosis; use ce-work when the change is already known. | 5 |
| ce-plan | You or the agent | Create structured plans for multi-step work, including software and non-software tasks. Use when asked to plan, break down implementation, plan from requirements, or deepen an existing plan; prefer ce-brainstorm for exploratory framing. | 13 |
| ce-polish | Only you | Polish a working feature through user-directed live browser feedback. Use when a functional feature needs focused UX refinement before shipping. | 3 |
| ce-pov | You or the agent | Judge a supplied subject against the project's evidence and constraints. Use when assessing an external-adoption question, a holistic take on a document, or a supplied approach set. Use for an oracle panel to consult other models and reconcile their opinions. Use ce-explain for understanding and ce-doc-review for findings review. | 10 |
| ce-product-pulse | Only you | Generate time-windowed product pulse reports from configured signals. | 2 |
| ce-promote | Only you | Draft launch or promotion copy for a shipped feature. | 2 |
| ce-proof | You or the agent | Publish, read, comment on, or edit markdown in Proof. Use for Proof links, sharing specs/plans/drafts, or publish handoffs from planning workflows; avoid proofread, math, evidence, or proof-of-concept meanings. | 3 |
| ce-prototype | You or the agent | Build a throwaway prototype to answer how something should work, feel, or read. Use when committing the wrong answer would be expensive to unravel and a cheap sketch cannot settle it. Not a rough visual probe during brainstorming, not for deciding what to build, not polishing a feature that already works, not implementing the real thing. | 4 |
| ce-resolve-pr-feedback | You or the agent | Judge PR feedback centrally, apply valid fixes, and complete review conversations with publication verified. Use when addressing feedback already left on a PR, preparing local fixes for a caller to publish, or completing a saved feedback batch. Use ce-code-review for reviewing code before feedback exists. | 6 |
| ce-retune | Only you | Retune a skill corpus for a new model, measurement-first: mine the run archive for a baseline, establish a noise floor, audit the corpus adversarially, then cut in measured passes until a pre-registered bar clears. Requires a benchmark harness that can A/B two builds of the corpus; refuses without one. | 0 |
| ce-riffrec-feedback-analysis | You or the agent | Analyze recorded product feedback into evidence for bugs and requirements. Use when a Riffrec capture or other screen, voice, or notes artifact needs interpretation. Use for Riffrec setup, capture, or sharing help when no recording exists yet. | 3 |
| ce-setup | Only you | Check Compound Engineering health and repo-local config, scaffold a Compound Pack with `pack:<id>`, or draft or refresh coding standards and review docs with `standards`. | 5 |
| ce-simplify-code | You or the agent | Simplify code for clarity, reuse, quality, and efficiency while preserving behavior, and prune tests and comments that carry no value. Use after implementation and before review, or to audit tests or comments in any part of a repo; use ce-debug for bugs. | 2 |
| ce-strategy | You or the agent | Create or update STRATEGY.md. Use when starting a product, adding a strategy doc, or changing direction or roadmap. | 5 |
| ce-sweep | Only you | Sweep configured feedback sources (Slack, GitHub Issues; email experimental) for new items: acknowledge at source, analyze recordings, verify fixes merged to main, and emit an `lfg`-ready plan. First run sets up sources; supports mode:non-interactive for scheduled runs. | 4 |
| ce-test-browser | You or the agent | Run browser tests for pages affected by the current branch or PR. Use when asked to run or check browser tests for the current change. | 2 |
| ce-test-xcode | Only you | Test iOS apps in a simulator with XcodeBuildMCP. Use when iOS changes need simulator evidence before handoff. | 1 |
| ce-work | You or the agent | Execute a plan or concrete work prompt end-to-end. Use when implementing from a plan document, a spec path, or a clear build request; use ce-debug for open-ended bugs. Use when an outer orchestrator needs implementation and local verification only, without the shipping tail. | 10 |
| ce-worktree | You or the agent | Set up isolated git worktrees — create a new branch for fresh work, or attach a worktree to an existing branch, PR, or commit. Use when starting isolated work or isolating an existing ref. | 3 |
| lfg | You or the agent | Take a request all the way to done, hands-off, through the right Compound Engineering skills. A code change ends as an open pull request, pushed without stopping. Use only when the user explicitly asks for autonomous end-to-end work or invokes lfg directly. Use ce-plan, ce-work, ce-debug, or ce-commit-push-pr for work the user reviews step by step. | 18 |
| wtf | Only you | Explain the last message, or a supplied file, link, or passage, in plain language a person follows on the first read. Use when the user invokes wtf because they did not follow something. Use ce-explain to investigate how or why code works, and ce-noslop to rewrite prose. | 2 |
From EveryInc/compound-engineering-plugin at 3449230, 2026-10-31. Skills read by DeepSeek V4.1 Flash on 2026-11-04. Summary and techniques by Claude Opus 5.5 on 2026-11-06.