result-to-claim

Use when experiments complete to judge what claims the results support, what they don't, and what evidence is still missing. A secondary Codex agent evaluates results against intended claims and routes to next action (pivot, supplement, or confirm). Use after experiments finish — before writing the paper or running ablations.

Quality

78%

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 ./skills/skills-codex/result-to-claim/SKILL.md

Quality

Content

73%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.

A well-structured, actionable workflow with strong sequencing and validation. The main headroom is in conciseness (heavy inline bash in Steps 1.5 and 5) and confirming that the referenced shared-references files actually exist in the bundle.

Suggestions

Move the long Step 5 research-wiki bash block and the Step 1.5 evidence pre-check script into a referenced file (e.g. shared-references/research-wiki-protocol.md) to reduce inline weight and improve conciseness.

Verify the shared-references files cited in the body (evidence-precheck.md, review-tracing.md, reviewer-routing.md, experiment-integrity.md) exist in the bundle; broken reference paths undermine progressive disclosure.

Tighten the prose around the spawn_agent template and routing tables — the per-verdict instructions repeat structure that could be condensed without losing actionability.

Dimension	Reasoning	Score
Conciseness	The body is mostly purposeful with no basic-concept padding, but Step 1.5 and the lengthy Step 5 research-wiki bash blocks are heavy inline content that could be tightened or moved to a reference.	3 / 5
Actionability	Provides concrete executable bash, wandb.Api() calls, a full spawn_agent template, and real fallback logic; minor gaps remain as placeholders are left for configurable values.	4 / 5
Workflow Clarity	Steps 1-5 are clearly sequenced with explicit validation checkpoints (Step 1.5 evidence pre-check, Step 3.5 integrity audit) and feedback loops (partial -> supplement -> re-run, fail-closed routing).	5 / 5
Progressive Disclosure	Good section structure with well-signaled one-level references, but the cited shared-references files are not present in the bundle and long inline bash (Step 5) could live in a referenced file.	4 / 5
	Total	16 / 20 Passed

Description

83%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.

A strong, clearly-scoped description that answers both 'what' and 'when' with concrete actions and natural triggers. Slight headroom remains on trigger synonym coverage and distinguishing itself from adjacent research-suite skills.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ('judge what claims the results support, what they don't, and what evidence is still missing' and 'routes to next action (pivot, supplement, or confirm)'), but stops short of the fully comprehensive coverage of a 5.	4 / 5
Completeness	Explicitly answers both what the skill does (judge/evaluate/route claims) and when to use it (after experiments finish, before writing the paper or running ablations) with concrete trigger phrases.	5 / 5
Trigger Term Quality	Natural triggers like 'Use when experiments complete' and 'Use after experiments finish — before writing the paper or running ablations' give good coverage, but a few common synonyms (e.g. 'results', 'evaluation done') are absent.	4 / 5
Distinctiveness Conflict Risk	The result-to-claim niche is distinct with specific triggers, but it sits in a suite of related research skills (idea-creator, ablation-planner, proof-checker) creating minor overlap risk.	4 / 5
	Total	17 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 suspicious	Warning

	Total	13 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/skills-codex/result-to-claim/SKILL.md
Commit: f4f20f9

Reviewed: 1 day 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.