rethink

Pursue the best possible version of a system, API, or abstraction, develop the vision through dialectic, and work backward to make it real. Use when the user asks for greenfield thinking, radical options, a clean break, first-principles design, or says the current abstraction feels wrong.

Quality

77%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/rethink/SKILL.md

Quality

Content

75%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is an exceptionally lean, well-sequenced piece of instruction writing with genuinely well-structured progressive disclosure. Its weakness is actionability: guidance is frequently directional rather than concrete, with no worked example to anchor what a good dialectic articulation or backward-derived plan looks like.

Suggestions

Add one compact worked example — e.g., a short before/after API sketch showing a vision articulated, then the backward-derived step order — to ground the abstract directives.

Specify the concrete artifacts the process should produce (a stated target design, a dependency-ordered change list, a verification checklist) so each phase has a checkable output.

Fix or clarify the dialectic link: either make it an in-bundle reference or describe the dialectic interaction inline so navigation does not depend on an external skill path.

Dimension	Reasoning	Score
Conciseness	Lean, dense, directive prose throughout — "Suspend inherited constraints long enough to discover the strongest shape", "A convenient intermediate shape must not quietly become the endpoint" — with no padding and no explanation of concepts Claude already knows. Every clause carries guidance; not 4 because there is nothing to trim.	5 / 5
Actionability	As an instruction-only skill, absence of code is not penalized, and some guidance is concrete ("Follow a caller, interaction, or data flow far enough to expose the decisions the model makes about ownership, state, policy, and lifetime"). But much is directional and abstract ("follow the idea all the way", "Let the destination determine the work") with no worked example or concrete artifact specification. Not 4 because the guidance stops short of executable specificity.	3 / 5
Workflow Clarity	A clear three-phase sequence — Pursue the vision, Work backward from the destination, Close against the vision — with a validation close ("Judge the result against the destination", "Exercise its concrete use and verify the changed contracts"). Not 5 because there is no explicit error-recovery/feedback loop describing what to do when validation reveals drift.	4 / 5
Progressive Disclosure	Three well-signaled one-level-deep references, all verified to exist (references/greenfield.md, references/radical-options.md, references/clean-breaks.md), with a body appropriately sized and split. Not 5 because the [dialectic](../../skills/dialectic/SKILL.md) link points outside the bundle to a path that does not resolve in the evaluation context, leaving a potentially broken navigation hop.	4 / 5
	Total	16 / 20 Passed

Description

70%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description cleanly answers both what the skill does and when to use it, with unusually good trigger phrasing including paraphrase triggers. Its main weakness is an abstract "what" — the actions are directional rather than concrete, which also limits specificity.

Suggestions

Make the "what" more operationally concrete, e.g., "articulates a target design through back-and-forth critique, then derives the implementation and migration plan from it".

Add natural synonyms users would say, such as "redesign", "rearchitect", or "rebuild from scratch", to broaden trigger coverage.

Dimension	Reasoning	Score
Specificity	Names the domain ("a system, API, or abstraction") and three actions ("Pursue the best possible version", "develop the vision through dialectic", "work backward to make it real"), but the actions are process-level and abstract rather than concrete operations. Not 4 because the verbs lack the concrete objects seen in anchor 4 (e.g., "fills forms, converts pages to images").	3 / 5
Completeness	Both what and when are explicitly present: the what ("Pursue the best possible version... develop the vision through dialectic, and work backward to make it real") and an explicit "Use when..." clause with multiple trigger phrases. Not 5 because the "what" remains abstract and could state the concrete outputs (e.g., a target design plus a derived implementation plan).	4 / 5
Trigger Term Quality	"greenfield thinking, radical options, a clean break, first-principles design, or says the current abstraction feels wrong" gives good natural-phrase coverage including a paraphrase trigger ("feels wrong"). Not 5 because common synonyms users would actually say — "redesign", "rearchitect", "rebuild from scratch" — are missing.	4 / 5
Distinctiveness Conflict Risk	A clear niche (greenfield/first-principles rethinking of abstractions) with distinctive trigger phrases like "a clean break" and "first-principles design". Not 5 because there is minor overlap risk with general design, planning, or refactoring skills whose requests can also involve "the current abstraction feels wrong".	4 / 5
	Total	15 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 1 suspicious	Warning

	Total	15 / 16 Passed

Repository: EpicenterHQ/epicenter
Path: .agents/skills/rethink/SKILL.md
Commit: f9441c8

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.