agentleFS
System-Error-WorldwideSystem-Error-Worldwide/the-loop/.agents/skills/strategize/SKILL.md

strategize

Skill0 starsChanged 3 months agoMITView on GitHub

Turn a stated intent into the smallest bounded problem, outcome, scope and testable success gate before implementation begins. Use when an idea, request or detected issue needs a strategy packet before specifications or implementation.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use for the first lifecycle stage, or when scope, objective or the definition of success has materially changed.

    line 19
  2. Very specific:

    run ID, stage strategize, track and declared asset.

    line 46
  3. Distinctive:

    Ask: does this solve the stated problem with the smallest sufficient scope, and could a different implementer satisfy the wording while …

    line 68

“Common” is measured: kept by at least 6 in 10 of the 3 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Strategize

Purpose

Provide the complete bundled fallback for lifecycle.strategize. Convert intent into a decision brief that is bounded enough to specify and verify. Do not draft implementation or hide a strategic fork.

Use when

Use for the first lifecycle stage, or when scope, objective or the definition of success has materially changed. An installed specialist may replace this fallback only through a recorded qualified route.

Required inputs

  • Validated run ID, declared asset, immutable code or noncode track and current stage.
  • User intent, known constraints, supplied sources and relevant prior decisions.
  • Frozen mission budget, active authority reference and exact mutation boundary.
  • Current issues and evidence references, if this is a revised strategy.
  • A requested capability and fallback route decision.

Reject the invocation when owner, lease, authority, asset, track or state cannot be established. Read-only analysis may continue only if the result is explicitly an unverified halt packet rather than a stage pass.

Procedure

  1. Restate the observed problem without adding unconfirmed facts.
  2. Define the desired outcome and the user or operator who benefits.
  3. Draw explicit in-scope, out-of-scope and later boundaries around one declared asset.
  4. List assumptions as confirmed, safely defaulted with a reversal condition, or unresolved human gates.
  5. Identify dependencies, authority needs, external effects, cost, privacy, legal and security risks.
  6. State success measures and turn each into a falsifiable done gate.
  7. Choose the smallest sufficient next unit: specification, one build slice, a non-code drafting unit, or a truthful halt.
  8. Record the bundled fallback as provider source and preserve every installed-candidate rejection reason.

Output contract

Return:

  • run ID, stage strategize, track and declared asset;
  • bounded problem statement and desired outcome;
  • in-scope, out-of-scope and later lists;
  • assumptions, constraints, risks and unresolved decisions;
  • success measures and exact done gate;
  • produced artifacts plus evidence and issue references;
  • resulting gate state and recommended next stage or halt reason.

Evidence gate

Pass only when each material assumption has a confirmation, safe default with reversal condition, or named human gate, and each success measure has a verification method. Persist passing evidence before advancing. Missing or stale required evidence is UNVERIFIED, never passed.

Track requirements

For the code track, include repository identity, worktree state, intended isolation, live-state coverage and executable acceptance behavior. For the non-code track, include audience, format, factuality standard, source threshold, review method and publication boundary.

Safety and authority

Strategy may recommend outward work but cannot perform it. Expansion to another asset, audience, integration or materially broader success condition requires a new authority decision. Every side effect remains behind runtime preflight, durable intent and an unexpired matching lease.

Self-refutation

Ask: does this solve the stated problem with the smallest sufficient scope, and could a different implementer satisfy the wording while producing the wrong outcome? Tighten any ambiguous gate before passing.

Halt conditions

Halt on a strategic fork that changes the asset, authority, cost, legal exposure, privacy exposure or success definition; missing required source state; conflicting constraints; exhausted budget; detected stop; lost ownership; or an unavailable verification method. Include the last completed action, evidence gap and exact decision needed to continue.

Core execution does not require network access or the source checkout. The links below are optional public documentation.

References

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.