New: 97,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Measured Optimization Loop

by EveryInc in EveryInc/compound-engineering-plugin

Optimizes a named target with a measured loop, attributing a workload's cost or scoring variants and keeping winners, on a dedicated branch with a disk log.

MITAuto-check passedDevelopment

Install Measured Optimization Loop

skills CLI
$ npx skills add EveryInc/compound-engineering-plugin --skill ce-optimize -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install EveryInc/compound-engineering-plugin ce-optimize --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/EveryInc/compound-engineering-plugin.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/ce-optimize .claude/skills/ce-optimize && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
ce-optimize
GitHub stars
25k
Token cost
~2k tokens
SKILL.md length
1,167 words
Files
20 (incl. scripts, references)
Skills in repo
37
Repo updated
First seen
Licence
MIT

At a glance

Optimizes a named target with a measured loop, attributing a workload's cost or scoring variants and keeping winners, on a dedicated branch with a disk log.

  • Reducing the runtime or cost of a specific working workload when the fix is not obvious
  • SKILL.md covers Interaction Method, Artifact Root, Persistence Discipline and The phases
  • Runs JavaScript and Shell scripts from its folder; calls git
  • Scoring several variants of a prompt, config or approach and keeping the winner

What it does

Use it when a working system's metric should improve and the winning change is not yet known. The agent either attributes the cost of a named workload before searching for implementations, or scores a space of variants and keeps the best, without needing a profile. Confirmed improvements end up as keep or revert commits on an `optimize/` branch named after the spec, with a log written to disk, and you take over at wrap-up.

Invoking the skill authorizes reading the repo, building the measurement harness and, after an approval gate in Phase 1, isolated experiments. The agent asks when spend is uncapped, when a new dependency appears, when wrap-up would push or open a PR, or when you must choose among post-completion options. It reports in ordinary task language, treats unknown duration or cost as unknown, and counts a step as done only after it ran. The folder ships spec and experiment-log schemas, example specs, prompt templates, `decide.mjs` and `experiment-worktree.sh`; ce-debug and ce-work cover diagnosis and known changes.

When your agent uses it

  • Reducing the runtime or cost of a specific working workload when the fix is not obvious
  • Scoring several variants of a prompt, config or approach and keeping the winner
  • Running experiments in isolated worktrees with keep-or-revert commits

Example prompts

  • “Speed up the nightly report job. Work out where the time goes before trying any fixes.”
  • “Try variants of the summarization prompt, score each against the judge spec and keep the best.”
  • “Cut our API cost per request with a measured loop and log every experiment.”

Requirements

  • A git repository where the agent can create an optimize branch and worktrees
  • A measurable metric or judge for the target

What it can do on your machine

Read from SKILL.md and the folder at commit cef001f. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 2 files in scripts/ (JavaScript and Shell, from the files we listed), which the agent can run.

    Shell commands in SKILL.md call:

    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Measured Optimization Loop loads about 2k tokens when it runs, and up to ~38k if it reads all its reference files. Until then it costs about 75 tokens; SKILL.md has 1,167 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~75
When it runs · the whole SKILL.md, loaded when a task matches
~2k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~38k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from EveryInc/compound-engineering-plugin at commit cef001f, republished under its MIT licence (© EveryInc). 1,167 words, ~1,970 tokens.

Download SKILL.mdSave it as .claude/skills/ce-optimize/SKILL.md (or your agent's skills folder). This skill also uses 19 other files; get the full folder from GitHub.
name
ce-optimize
description
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.
argument-hint
[path to optimization spec YAML, or describe the optimization goal]

Optimize a measurable target

Outcome: confirmed improvements to the named target live on an optimize/<spec-name> branch, with a disk log. The user takes over at wrap-up.

Intent: the next action is the cheapest step that would change what gets implemented. Attribute the cost of a named workload before searching implementations. Search and keep a scored variant space without requiring a profile.

Done when: a stopping criterion was met, every declared required target is met or another stop was met first, the final state is written and verified on disk, and the user has been given the post-completion options. If the run instead stopped at a check it could not pass, say what blocked it.

Invoking this skill authorizes reading the repo, building the harness, and (after the Phase 1 approval gate) isolated experiments and keep/revert commits on optimize/<spec-name>. Ask when spend is uncapped, when a new dependency appears, when wrap-up would push or open a PR, or when only the user can choose among the post-completion options. Do not ask again to run the next experiment inside those limits.

Independent calls and dispatches that do not depend on each other go in one response. Serialize only real dependencies.

Report findings, user decisions, blockers, and results. During longer work, give occasional updates on what was learned and what remains. Routine preparation and phase or batch transitions need no separate announcement. Keep accounting in the log and final recap unless it affects a current decision.

Explain the target, evidence, and decision in ordinary task language. Workflow labels (such as 'harness' or 'parallel readiness') belong in artifacts unless the user asks about those mechanics. State unknown duration or cost as unknown; caps are limits, not forecasts.

A step is done only after it ran. Describing a measurement, dispatch, or checkpoint is not doing it. Do not end a turn while in-scope work remains merely described.

Interaction Method

Use the host's blocking question tool already in the current tool list (match by capability, not by a host-specific name). Presence in the current tool list is proof the tool exists; never call a user-facing question tool to discover whether it exists. If a matching tool is listed but unloaded, use the host's tool-discovery primitive to load that capability: do not search for another host's tool name. Fall back to numbered options on the host's chat surface only when no such tool is in the list or a real question call errors. Never skip the question silently.

Artifact Root

Resolve <root> the first time you compose a path under it. Reading learnings under <root>/solutions/ counts as composing one. Give any subagent the resolved path, not the config.

<!-- ce-docs-root:start -->

Resolve the CE artifact root <root> before composing any artifact path.

  • Read docs_root from <repo-root>/.compound-engineering/config.yaml only (<repo-root> = git rev-parse --show-toplevel). Do not read it from config.local.yaml. Unset -> <root> is docs, exactly as before.
  • Validate a set value: a repo-relative directory whose real, symlink-resolved path stays inside the repo and is neither the repo root nor under .git/. Otherwise stop with an error naming docs_root and the value -- never fall back to docs.
  • Use <root> as the sole artifact location: create it if absent, compose each path as <root>/<subdir> with this skill's own subdirectory, and never also read docs.
<!-- ce-docs-root:end -->

Persistence Discipline

The experiment log on disk is the source of truth. Write order is measure, write, verify, then show the user. Read references/persistence.md now for checkpoints CP-0 through CP-5, the file layout, and resume. The phases below mark where each checkpoint falls.

Show full SKILL.md (585 more words)Show less

The phases

Four phases run in order. Each one names the reference it cannot start without. A fresh run skips none of them: a harder optimization spends longer in a phase, it does not run fewer phases.

A resume is not a fresh run. On a resume, re-enter Phase 0 only far enough to detect the run and to recover any result.yaml markers the log is missing. Then continue from the phase the log records: skip the work the log proves finished, and re-enter any approval check it does not. A checkpoint proves the work that produced it, never a user decision: the log holds no record of approval, so a resume that has not seen the user approve presents the Phase 1 gate again.

Phase 0: Setup. The input is a goal, or a path to a spec YAML. It comes from the user or from a calling skill. If neither supplied one, ask: "What would you like to optimize? Describe the goal, or provide a path to an optimization spec YAML file." Load or build the spec and save it (CP-0): read references/spec.md. Then search prior learnings, detect run identity, and create the branch and scratch space. Read references/measurement.md for the rest of Phase 0 and Phase 1.

Phase 1: Measurement scaffolding. Build or validate the harness, write the baseline (CP-1), probe parallelism, check the worktree budget. Two gates stop the run:

  • Clean-tree gate. Do not continue while any file in scope.mutable or scope.immutable has uncommitted changes. The reference defines the check and what to ask for.
  • User approval gate. Present what Phase 1 assembled; the reference lists what to include. Say plainly which spend is capped. stopping.max_total_cost_usd caps the whole run, while the judge cap counts only scoring. With neither set, spend is uncapped apart from the iteration and hour limits. Offer proceed, fix issues, and adjust spec. Adjusting the spec is only available while the log holds nothing derived from it (no hypothesis backlog and no experiments) and it sends the run back through Phase 1 so the baseline matches the new spec. Once anything derived from the spec is on file, the spec is fixed for the run. Do not enter Phase 2 until the user explicitly approves. Then re-read the spec and baseline from disk.

Phase 2: Hypothesis generation. Analyze the current approach, rank the hypotheses, record the backlog (CP-2). Do not dispatch an implementation experiment while a cheaper locating measurement would change keep or skip. Read references/loop.md for this phase and Phase 3. One gate: dependency pre-approval. Collect every new dependency across all hypotheses and present the full list for bulk approval. A dependency the user does not approve stays in the backlog, is skipped in batch selection, and comes back at wrap-up.

Phase 3: Optimization loop. Select a batch, dispatch experiments, persist each result as it lands (CP-3), evaluate with scripts/decide.mjs, update state and the digest (CP-4), then check whether to stop. Stop as soon as any one of seven criteria holds: every declared required target is met, max iterations, max hours, a spend cap reached, plateau, a user interrupt, or no runnable hypothesis left. references/loop.md states each one exactly. Otherwise start the next batch.

Phase 4: Wrap-up. Read references/wrap-up.md for the deferred hypotheses, the summary, what is preserved, cleanup, and the post-completion options to present. CP-5 marks the log final. Write it only after the user picks an option that does not return to Phase 3. Two options do return: Continue, and approving a deferred dependency.

© EveryInc, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 19 other files (scripts, references) in skills/ce-optimize of EveryInc/compound-engineering-plugin.

  • SKILL.md
  • references/agents/learnings-researcher.md
  • references/agents/repo-research-analyst.md
  • references/example-expensive-benchmark-spec.yaml
  • references/example-hard-spec.yaml
  • references/example-judge-spec.yaml
  • references/experiment-log-schema.yaml
  • references/experiment-prompt-template.md
  • references/judge-prompt-template.md
  • references/loop.md
  • references/measurement.md
  • references/optimize-spec-schema.yaml
  • references/persistence.md
  • references/spec.md
  • references/usage-guide.md
  • references/wrap-up.md
  • scripts/decide.mjs
  • scripts/experiment-worktree.sh
  • … and 2 more

Open the folder on GitHubat commit cef001f

Compare with similar skills

Measured Optimization Loop next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Measured Optimization Loop compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Measured Optimization Loop this skillEveryInc/compound-engineering-plugin25k—~2kAutomated safety check: PassMIT
Self-Improvement Tournament Loopzereight/gitlab-mcp2k1 repos~1.8kAutomated safety check: WarnMIT
Hiverllm-org/hive216—~2.1kAutomated safety check: PassApache-2.0
Session Profilertamdogood/builder-essential-skills218—~1.5kAutomated safety check: PassMIT
Build Dotnet Agentnewrelic/newrelic-dotnet-agent117—~871Automated safety check: PassApache-2.0
Effective Harnessesliangdabiao/exa-research-mcp-skill109—~1.5kAutomated safety check: PassNone

Similar skills

  • Self-Improvement Tournament Loop

    zereight/gitlab-mcp

    Runs an autonomous evolutionary loop that improves a codebase against a measurable benchmark, using agent roles, tournament selection and recorded history until a stop condition.

    2k GitHub starsUsed in 1 repo~1.8k tokens
    DevelopmentAuto-check: warnings

  • Hive

    rllm-org/hive

    Run the hive experiment loop — autonomous iteration on a shared task.

    216 GitHub stars~2.1k tokensUpdated 5 mo ago
    DevelopmentAuto-check passed

  • Session Profiler

    tamdogood/builder-essential-skills

    Profile and debug Hermes sessions from their JSONL transcripts.

    218 GitHub stars~1.5k tokensUpdated 1 mo ago
    DevelopmentAuto-check passed

  • Build Dotnet Agent

    newrelic/newrelic-dotnet-agent

    Build the New Relic .NET agent locally and validate that code changes compile.

    117 GitHub stars~871 tokensUpdated today
    DevelopmentAuto-check passed

  • Effective Harnesses

    liangdabiao/exa-research-mcp-skill

    Long-running agent project harness for Codex, OpenClaw, Claude Code, and other coding agents.

    109 GitHub stars~1.5k tokensUpdated 1 mo ago
    DevelopmentAuto-check passed

  • PR Greenlight

    UniClipboard/UniClipboard

    Agent Loop that runs local pre-flight CI checks, auto-fixes issues, creates/pushes the PR, monitors CI, and loops until all checks pass.

    1.8k GitHub stars~2.8k tokensUpdated today
    DevelopmentAuto-check passed

More from EveryInc/compound-engineering-plugin

All 37 skills in this repo

  • Compound Learning Writer

    EveryInc/compound-engineering-plugin

    Records one solved and verified problem as a durable learning in the repository, but only when the reasoning is not already clear from the final code, tests or docs.

    25k GitHub stars~2k tokensUpdated today
    Auto-check passed

  • Compound Learnings Refresh

    EveryInc/compound-engineering-plugin

    Audits a repo's stored learnings against the current codebase, fixes stale, overlapping or superseded docs and reports on every document.

    25k GitHub stars~2k tokensUpdated today
    Auto-check passed

  • Compound Engineering Prototype

    EveryInc/compound-engineering-plugin

    Builds a throwaway prototype at just the fidelity needed to settle a specific how-it-should-work-or-feel question, before committing to an approach other work will treat as fixed.

    25k GitHub stars~1.9k tokensUpdated today
    Auto-check passed

  • Compound Engineering Setup

    EveryInc/compound-engineering-plugin

    Checks Compound Engineering plugin health and repo-local config, or scaffolds a Compound Pack when you ask for one by id.

    25k GitHub stars~1.9k tokensUpdated today
    Auto-check passed

  • PR Babysitter

    EveryInc/compound-engineering-plugin

    Watches an open GitHub pull request over time, routing review comments and CI failures to other skills until the PR is ready to merge.

    25k GitHub stars~2k tokensUpdated today
    Auto-check passed

  • CE Brainstorm

    EveryInc/compound-engineering-plugin

    Turns a vague or ambitious feature idea into a requirements-only plan through dialogue with you, sized to the work, before any code is written.

    25k GitHub stars~1.9k tokensUpdated today
    Auto-check passed

Categories

Questions about Measured Optimization Loop

What does Measured Optimization Loop do?

Optimizes a named target with a measured loop, attributing a workload's cost or scoring variants and keeping winners, on a dedicated branch with a disk log. Use it when a working system's metric should improve and the winning change is not yet known. The agent either attributes the cost of a named workload before searching for implementations, or scores a space of variants and keeps the best, without needing a profile.

When should I use Measured Optimization Loop?

Measured Optimization Loop fits situations like: reducing the runtime or cost of a specific working workload when the fix is not obvious; scoring several variants of a prompt, config or approach and keeping the winner; running experiments in isolated worktrees with keep-or-revert commits.

How do I install Measured Optimization Loop in Claude Code?

Run `npx skills add EveryInc/compound-engineering-plugin --skill ce-optimize -a claude-code`. Or copy the skill folder (skills/ce-optimize in EveryInc/compound-engineering-plugin) into .claude/skills/ce-optimize in your project. Claude Code loads it when a task matches its description.

How do I install Measured Optimization Loop in Codex?

Run `npx skills add EveryInc/compound-engineering-plugin --skill ce-optimize -a codex`. Or copy the skill folder (skills/ce-optimize in EveryInc/compound-engineering-plugin) into .agents/skills/ce-optimize in your project. Codex loads it when a task matches its description.

Can I use Measured Optimization Loop in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add EveryInc/compound-engineering-plugin --skill ce-optimize -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/ce-optimize, .gemini/skills/ce-optimize, .github/skills/ce-optimize and .opencode/skills/ce-optimize in your project.

What does Measured Optimization Loop need to run?

Going by SKILL.md and its folder, Measured Optimization Loop needs JavaScript and a shell for the scripts in its folder and the command-line tools its instructions call (git). Our summary lists: A git repository where the agent can create an optimize branch and worktrees; A measurable metric or judge for the target.

Does Measured Optimization Loop access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Measured Optimization Loop safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Measured Optimization Loop use?

Measured Optimization Loop is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Measured Optimization Loop use?

About 2k tokens (SKILL.md is roughly 7.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 36k tokens, read only when the agent opens those files.

What are the alternatives to Measured Optimization Loop?

Skills that share tags, products or a category with Measured Optimization Loop: Self-Improvement Tournament Loop (zereight/gitlab-mcp, 2k stars), Hive (rllm-org/hive, 216 stars), Session Profiler (tamdogood/builder-essential-skills, 218 stars) and Build Dotnet Agent (newrelic/newrelic-dotnet-agent, 117 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Measured Optimization Loop?

EveryInc (a GitHub organization) maintains it in EveryInc/compound-engineering-plugin, which has 25,412 GitHub stars. The repository holds 37 skills in this directory. The repository was last updated on October 7, 2026.

Source: EveryInc/compound-engineering-plugin on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.