experiment-bridge

Workflow 1.5: Bridge between idea discovery and auto review. Reads EXPERIMENT_PLAN.md, implements experiment code, deploys to GPU, collects initial results. Use when user says "实现实验", "implement experiments", "bridge", "从计划到跑实验", "deploy the plan", or has an experiment plan ready to execute.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

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.

The body is a well-sequenced, validation-rich workflow with strong actionability, though it carries verbosity in the embedded review prompt and placeholder-heavy command examples. Progressive disclosure is good but relies on external shared-references rather than a local bundle.

Suggestions

Trim or externalize the long spawn_agent code-review prompt in Phase 2.5 to a reference file, keeping only the invocation and key checks inline.

Replace placeholder command blocks (e.g. '/run-experiment [experiment commands]') with at least one concrete worked example so guidance is copy-paste ready.

Consolidate the repeated ground-truth-evaluation guidance into a single stated rule to reduce token redundancy across Phases 2, 2.5, and Key Rules.

Dimension	Reasoning	Score
Conciseness	Mostly efficient, but the multi-paragraph code-review spawn_agent prompt (Phase 2.5) and repeated restatements of the ground-truth rule pad the body with content that could be tightened or moved out.	3 / 5
Actionability	Provides concrete commands, file paths, and template blocks, but several examples remain placeholder-laden (e.g. '/run-experiment [experiment commands]', '[N experiments, ~X GPU-hours]') rather than copy-paste ready.	4 / 5
Workflow Clarity	Six clearly sequenced phases with explicit validation checkpoints (sanity stage, fresh-agent code review, deploy checkpoint) and concrete failure-feedback loops ('read the traceback first', 'discard and reimplement cleanly').	5 / 5
Progressive Disclosure	Well-organized sections with clear navigation to shared-references files (output-versioning, external-cadence, review-tracing), but no bundle files exist and references point to sibling shared-references rather than one-level-deep local files.	4 / 5
	Total	16 / 20 Passed

Description

92%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 is specific, complete, and distinctive, clearly stating what the skill does and when to invoke it with concrete multilingual trigger phrases. Minor keyword synonym coverage gaps keep trigger quality just below the top anchor.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Reads EXPERIMENT_PLAN.md, implements experiment code, deploys to GPU, collects initial results' — giving comprehensive coverage of the skill's capabilities.	5 / 5
Completeness	Explicitly answers both what (reads plan, implements, deploys, collects results) and when ('Use when user says ...'), with concrete trigger phrases.	5 / 5
Trigger Term Quality	Strong keyword coverage across English ('implement experiments', 'deploy the plan') and Chinese ('实现实验', '从计划到跑实验') plus a state trigger, but a few natural synonyms ('run experiments', 'execute plan') are absent.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche as 'Workflow 1.5' bridging idea discovery and auto review, with distinct bilingual triggers and minimal overlap with adjacent skills.	5 / 5
	Total	19 / 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: 4 suspicious	Warning

	Total	15 / 16 Passed

Repository: wanshuiyin/Auto-claude-code-research-in-sleep
Path: skills/skills-codex/experiment-bridge/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.