SKILLEMALL.ai
RU Sign in

AF competitive-intelligence-deep-research

Run a multi-source, agentic deep-research loop on a single competitor and produce an evidence-backed narrative dossier with citations. Scaffolds queries across SEC EDGAR, the company's own site, press releases, Wikipedia REST, Crossref, and public reputation signal (G2 / TrustPilot / Reddit threads), dedupes, screens snippets for relevance, extracts dated facts, and synthesizes dossier.md with an audit log. Trigger when the user asks for "competitive intelligence on", "deep dive on", "competitor dossier", "build a profile on", or "research the competition".

genspark Agent Skills author: Genspark 1 file body ≈ 2 303 tokens Open the sourcegenspark.ai↗ analyzed 2 d ago

As a process F 65/100 · Will not run — References files that are not bundled: scripts/search_sources.py, scripts/format_dossier.py

AnalyzerAI and agentsResearchOperations and projectstype and topics are labelled automatically from the skill text
Runs in: Genspark 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) OpenClaw Goose OpenCode
JSON
Technical rating
A
93/100
safety, quality, tests
Safety 60%
100
Quality 40%
83
Run on models
none yet
Process rating
F
65/100
Will not run
References files that are not bundled: scripts/search_sources.py, scripts/format_dossier.py
Tools and files w 18
0
Failures and branches w 10
50
Result and completion w 14
60
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: 1. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning missing-ref reference to a missing file: scripts/search_sources.py
  • warning missing-ref reference to a missing file: scripts/format_dossier.py

Process rating: all ten parameters 65/100

Will not run. References files that are not bundled: scripts/search_sources.py, scripts/format_dossier.py
  • 0Tools and files. 2 referenced file(s) missing: scripts/search_sources.py, scripts/format_dossier.py
  • 50Failures and branches. 0 branches, has a failure section
  • 60Result and completion. Output format stated, no completion criterion
  • 70When it triggers. States when to use, but not when not to
  • 70Inputs and preconditions. Inputs and preconditions are listed
  • 100Steps. 45 steps
  • 100Consistency. Name and required fields are in place
  • 100Execution cost. Instruction body is 2303 tokens
  • 100Running it twice. Mutating operations check current state
  • 100Progress reporting. Reports progress
  • low The response is described with custom markup (3 tags): a typed call is more reliable

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

  • +4Description does not say when NOT to use the skill (false activations)
  • +1No license
  • +2Single-language instructions
  • +5Description quotes 5 example trigger phrases
  • +3Description length 563: enough signal without eating the budget
  • +4Structure: 8 headings
  • +3Step-by-step instructions: 45 items
  • +3Output format is stated explicitly
  • +4Has examples (2 code blocks)

Quality base 70; lint remarks subtract, signals add up to 100. Result: 83.