SKILLEMALL.ai

AF codex-deepsearch

Use when Codex must run or review deep, source-backed research as a Codex-native workflow: local/web evidence separation, tool-limit disclosure, source and claim ledgers, review gates, spec/roadmap/implementation readiness, /goal checkpointing, privacy-safe repo handling, and optional codex-with-cc-plus or Claude Code worker delegation. Use for evidence-heavy research, current product/API/tool/law/market questions, architecture discovery before implementation, and research that may become a plan, spec, roadmap, PRD, or engineering input.

ClawHub Agent Skills author: 少卿 v0.1.0 MIT-0 2 files body ≈ 1 359 tokens Open the sourceclawhub.ai analyzed 2 d ago

Use when Codex must run or review deep, source-backed research as a Codex-native workflow: local/web evidence separation, tool-limit disclosure, source and…

As a process F 56/100 · Will not run — References files that are not bundled: references/source-and-claim-ledgers.md, references/review-gate.md, references/codex-runtime-model.md

AnalyzerSoftware developmenttype and topics are labelled automatically from the skill text
JSON
Technical rating
A
90/100
safety, quality, tests
Safety 60%
100
Quality 40%
75
Run on models
none yet
Process rating
F
56/100
Will not run
References files that are not bundled: references/source-and-claim-ledgers.md, references/review-gate.md, references/codex-runtime-model.md
Tools and files w 18
0
Inputs and preconditions w 11
0
Progress reporting w 2
0
the three weakest of ten parameters · all ten

How to improve

  1. The text references files that are not there: add them or drop the references.
For the model run — optional
  • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
  • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

Guard findings · 0

✓ No critical or high findings

Files scanned: 2. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning missing-ref reference to a missing file: references/source-and-claim-ledgers.md
  • warning missing-ref reference to a missing file: references/review-gate.md
  • warning missing-ref reference to a missing file: references/codex-runtime-model.md
  • warning missing-ref reference to a missing file: references/research-workflow.md
  • warning missing-ref reference to a missing file: templates/RESEARCH_DESIGN.md
  • warning missing-ref reference to a missing file: templates/SOURCE_LEDGER.md
  • warning missing-ref reference to a missing file: templates/CLAIM_LEDGER.md
  • warning missing-ref reference to a missing file: templates/DEEPSEARCH_REPORT.md
  • warning missing-ref reference to a missing file: templates/REVIEW_GATE.md
  • warning missing-ref reference to a missing file: references/codex-tooling-rules.md
  • warning missing-ref reference to a missing file: references/privacy-and-repo-safety.md
  • warning missing-ref reference to a missing file: references/delegation-with-codex-with-cc.md

Process rating: all ten parameters 56/100

Will not run. References files that are not bundled: references/source-and-claim-ledgers.md, references/review-gate.md, references/codex-runtime-model.md
  • 0Tools and files. 12 referenced file(s) missing: references/source-and-claim-ledgers.md, references/review-gate.md, references/codex-runtime-model.md
  • 0Inputs and preconditions. Does not say what the process needs to start
  • 0Progress reporting. Says nothing while it works
  • 30Running it twice. 2 mutating operations with no state check
  • 50When it triggers. No condition that starts the skill
  • 55Failures and branches. 1 branches
  • 100Steps. 38 steps
  • 100Result and completion. Output format and completion criterion are stated
  • 100Consistency. Name and required fields are in place
  • 100Execution cost. Instruction body is 1359 tokens
  • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing

Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

Quality signals

  • +5Description has no quoted example phrases that should trigger the skill
  • +4Description does not say when NOT to use the skill (false activations)
  • +1No license
  • +2Single-language instructions
  • +3Description length 543: enough signal without eating the budget
  • +4Structure: 8 headings
  • +3Step-by-step instructions: 38 items
  • +3Output format is stated explicitly
  • +4Has examples (1 code blocks)

Quality base 70; lint remarks subtract, signals add up to 100. Result: 75.

External checks

ClawHub: clean
The flagged instruction appears to be a user-directed editing workflow, not hidden self-modification or unsafe persistence.
LLM: benign (medium) · VirusTotal: · 26 Jun 2026