SKILLEMALL.ai
GitHub ↗ RU Sign in

BF deep-research

Async deep research via Gemini Interactions API (no Gemini CLI dependency). RAG-ground queries on local files (--context), preview costs (--dry-run), structured JSON output, adaptive polling. Universal skill for 30+ AI agents including Claude Code, Amp, Codex, and Gemini CLI.

modbender/skill-library-mcp Agent Skills author: modbender MIT 21 files body ≈ 3 554 tokens Open the sourcegithub.com↗ analyzed 3 d ago

Async deep research via Gemini Interactions API (no Gemini CLI dependency).

As a process F 47/100 · Will not run — References files that are not bundled: {baseDir}/AGENTS.md

IntegrationAI and agentsData and analyticstype and topics are labelled automatically from the skill text
Runs in: OpenClaw Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) Goose OpenCode
JSON
Technical rating
B
75/100
safety, quality, tests
Safety 60%
73
Quality 40%
77
Run on models
none yet
Process rating
F
47/100
Will not run
References files that are not bundled: {baseDir}/AGENTS.md
Tools and files w 18
0
Progress reporting w 2
0
When it triggers w 12
20
the three weakest of ten parameters · all ten

What is at stake

Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.

Dangerous commands medium severity

Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.

If you install

The skill contains commands that delete files, rewrite disks or execute code fetched from the network. The agent may run them without asking if it believes the instructions require it.

For the author

Replace destructive commands with safe equivalents that ask for confirmation, scope them to one folder, and stop piping curl into a shell: pin a version and a checksum.

Exfiltration medium severity

Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.

If you install

The instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".

For the author

If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.

How to improve

  1. Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
  2. 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 · 19

✓ No critical or high findings

Medium and low: 19
  • medium Dangerous commands cmd-pipe-to-shell CHANGELOG.md:65
    Downloads and executes remote code from an unrecognised host (pipe to shell) (quoted — discussed, not commanded)
    - Removed all `curl | sh` install patterns from documentation (VirusTotal supply chain risk flag)
    quoted
  • medium Exfiltration intent-browser-credential-store scripts/upload.py:404
    Accesses a browser credential / cookie store (detector / deny-list definition)
    """Check upload operation status from local state."""
    detector
  • low Dangerous commands cmd-pipe-to-shell CLAUDE.md:15
    Downloads and executes remote code from an unrecognised host (pipe to shell) (negated — the text forbids it)
    6. **No curl|sh patterns**: Never use `curl ... | sh` install patterns in documentation. Link to official install docs pages instead.
    negated
  • low Dangerous commands cmd-pipe-to-shell CLAUDE.md:36
    Downloads and executes remote code from an unrecognised host (pipe to shell) (negated — the text forbids it)
    8. Verify no `curl|sh` patterns: `grep -r 'curl.*install.*sh' scripts/ *.md`
    negated
  • low Obfuscation obf-base64-blob docs/examples/event-sourcing-patterns.md:151
    Long base64-looking blob (test fixture / example file)
    1. [confluent.io](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…Gqw
    fixture
  • low Obfuscation obf-base64-blob docs/examples/event-sourcing-patterns.md:153
    Long base64-looking blob (test fixture / example file)
    3. [microsoft.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…9XH
    fixture
  • low Obfuscation obf-base64-blob docs/examples/event-sourcing-patterns.md:154
    Long base64-looking blob (test fixture / example file)
    4. [cqrs.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…WPY
    fixture
  • low Obfuscation obf-base64-blob docs/examples/event-sourcing-patterns.md:157
    Long base64-looking blob (test fixture / example file)
    7. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…Lb0
    fixture
  • low Obfuscation obf-base64-blob docs/examples/event-sourcing-patterns.md:160
    Long base64-looking blob (test fixture / example file)
    10. [medium.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…cRn
    fixture
  • low Obfuscation obf-base64-blob docs/examples/webassembly-state-of-art.md:125
    Long base64-looking blob (test fixture / example file)
    1. [reddit.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…NkO
    fixture
  • low Obfuscation obf-base64-blob docs/examples/webassembly-state-of-art.md:129
    Long base64-looking blob (test fixture / example file)
    5. [dev.to](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…dyw
    fixture
  • low Obfuscation obf-base64-blob docs/examples/webassembly-state-of-art.md:134
    Long base64-looking blob (test fixture / example file)
    10. [youtube.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…qhV
    fixture
  • low Obfuscation obf-base64-blob docs/examples/webassembly-state-of-art.md:135
    Long base64-looking blob (test fixture / example file)
    11. [dev.to](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…CZe
    fixture
  • low Obfuscation obf-base64-blob docs/examples/webassembly-state-of-art.md:142
    Long base64-looking blob (test fixture / example file)
    18. [bytecodealliance.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…idX
    fixture
  • low Obfuscation obf-base64-blob docs/examples/websocket-vs-sse-comparison.md:203
    Long base64-looking blob (test fixture / example file)
    2. [medium.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…Zes
    fixture
  • low Obfuscation obf-base64-blob docs/examples/websocket-vs-sse-comparison.md:208
    Long base64-looking blob (test fixture / example file)
    7. [stackoverflow.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…9BB
    fixture
  • low Obfuscation obf-base64-blob docs/examples/websocket-vs-sse-comparison.md:209
    Long base64-looking blob (test fixture / example file)
    8. [javascript.info](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…ou8
    fixture
  • low Obfuscation obf-base64-blob docs/examples/websocket-vs-sse-comparison.md:212
    Long base64-looking blob (test fixture / example file)
    11. [github.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…eK-
    fixture
  • low Obfuscation obf-base64-blob docs/examples/websocket-vs-sse-comparison.md:215
    Long base64-looking blob (test fixture / example file)
    14. [timeplus.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZI…Esn
    fixture

Files scanned: 20. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning description-no-when description does not say WHEN to use the skill (no "use when")
  • warning missing-ref reference to a missing file: {baseDir}/AGENTS.md

Process rating: all ten parameters 47/100

Will not run. References files that are not bundled: {baseDir}/AGENTS.md
  • 0Tools and files. 1 referenced file(s) missing: {baseDir}/AGENTS.md
  • 0Progress reporting. Says nothing while it works
  • 20When it triggers. No condition that starts the skill
  • 30Running it twice. 7 mutating operations with no state check
  • 40Consistency. Frontmatter name (deep-research) differs from the folder (agent-deep-research)
  • 50Failures and branches. 0 branches, has a failure section
  • 60Result and completion. Output format stated, no completion criterion
  • 70Inputs and preconditions. Inputs and preconditions are listed
  • 85Steps. 11 steps, 1 vague phrases
  • 100Execution cost. Instruction body is 3554 tokens
  • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
  • low 15 top-level sections: this looks like several domains in one skill

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)
  • +2Single-language instructions
  • +3Description length 276: enough signal without eating the budget
  • +4Structure: 29 headings
  • +3Step-by-step instructions: 11 items
  • +3Output format is stated explicitly
  • +4Has examples (19 code blocks)
  • +4Reference files are cited in the instructions (1 of 3)
  • +3All 5 scripts are documented
  • +1License stated

Quality base 70; lint remarks subtract, signals add up to 100. Result: 77.