agentsclimarketplace

Competlab funding watch

Skill competlab/competlab-ci-skills/skills/competlab-funding-watch

Competitive intelligence skills for AI agents — landscape analysis, competitor dossiers, AI visibility reports, weekly briefings, and battlecards. Powered by CompetLab. Works with Claude Code, Cursor, Codex, Gemini CLI.

Install
npx -y skills add competlab/competlab-ci-skills --skill competlab-funding-watch

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Surfaces recent funding events, ownership changes, ARR estimates, M&A activity, executive transitions, and category-adjacent capital pressure for each monitored competitor. Adapts output per-vendor based on funding mode (Public-company, PE-owned, Bootstrap, VC-stage, M&A-volatile). Critical for strategic context (capital intensity = competitor's ability to outspend/outscale you). Use when the user asks "funding posture", "competitor funding", "category capital pressure", "is competitor X well-funded", "ARR estimates", "recent rounds in [category]". Requires CompetLab MCP + Perplexity MCP + mcp__competlab__fetch_url for URL verification.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

5.8 KB, as published. Nobody here has run it

Funding & Capital Intelligence

You surface recent funding events + capital posture for each monitored competitor. Adapt output per the 5 funding modes — each mode requires different signals.

The 5 funding modes (classify each vendor first)

  1. Public-company mode — lead with market cap, earnings, guidance, share repurchases (e.g., "[$X]B market cap, [latest-quarter] [$Y]M revenue [+Z%] YoY, [$W]M share buyback authorized")
  2. PE-owned mode — lead with ownership history + no-new-rounds signal (e.g., "[Vendor] acquired by [PE firm] in [year]; no new rounds since")
  3. Bootstrap mode — lead with ARR estimates from public-aggregator sources (GetLatka, Prospeo, Growjo) (e.g., "$0 disclosed funding, ~$XM ARR estimate per GetLatka, recent M&A activity")
  4. VC-stage mode — lead with round history + valuation + lead investor (e.g., "[$X]M across [N] rounds, Series [letter] [$Y]M led by [investor] at [$Z]B valuation")
  5. M&A-volatile mode — three-way sequencing matters more than amount (e.g., "[vendor]'s 2025 sequence: [acquirer A] [$X]B announced [date] → regulatory block on IP carve-out → [acquirer B] [$Y]B reverse acquihire [date] → [acquirer C] acquired remaining IP [date]")

Workflow

Step 1: Identify competitors + classify

Pull list_projects + list_competitors. For each, infer mode from quick Perplexity check or known status.

Step 2: Per-competitor Perplexity research (with URL verification)

For each, one focused query:

"For [company name in {category}] in {current year}: give (1) funding posture + most recent round/M&A with dates, (2) ARR estimate if known, (3) headcount or recent hiring direction signals, (4) recent exec transitions, (5) any category-adjacent acquirer/investor activity. Cite specific sources."

Step 3: URL-verify every citation

Per PATTERN-url-verification.md: any Perplexity-returned URL gets probed via mcp__competlab__fetch_url with bodyNeeded:true, cleanHtml:true before surfacing in skill output. Drop hallucinated URLs. Validate cited amounts against source page where possible.

Common URL types: press releases (companies' own), Crunchbase (often 403-blocked), GetLatka (often accessible), TechCrunch + The SaaS News + sector trades, fund websites (Sprints Capital, Sequoia, etc.).

Step 4: Synthesize per-vendor + cross-competitor

Per vendor:

  • Funding mode (one of 5 above)
  • Most recent round/M&A: amount, date, lead investor, valuation if disclosed
  • Total funding to date
  • ARR estimate (with source/year)
  • Headcount + direction (growing/flat/shrinking with source)
  • Exec transitions in last 12 months
  • Notable signals (e.g., "8+ years funding-quiet" or "post-merger integration mode")

Cross-vendor:

  • Category funding pattern: bimodal (well-funded + bootstrapped) / converging / cooling
  • Recent M&A activity in category
  • Adjacent-VC pressure (vendors in tier-up category that competitors will eventually compete against)
  • Investor concentration (e.g., "Balderton led both Vendor X + Vendor Y rounds")

Output Structure

# Funding & Capital — [Category / Project]
> Generated [date] | X of N citations URL-verified

## Summary
[Category funding pattern + cross-cutting investor signals]

## Per-competitor
### [Competitor] — [Funding mode]
- Most recent: [amount + date + lead]
- Total: [$X]
- ARR: [if known, with year + source]
- Headcount: [if known, with trend]
- Notable: [posture signal — growth-mode, mature, stale, etc.]

## Cross-competitor patterns
[Bimodal/converging signals, M&A clusters, investor concentration]

Decision Questions

  1. "Is the category's adjacent-VC pressure (e.g., Brevo $500M raise next door) a threat or an acquisition opportunity for us?"
  2. "Are we benchmarking against the bootstrapped lean operator or the VC-stage growth bet — and is our current capital posture aligned?"

Error Handling

  • Perplexity returns "I don't have enough information" or thin/dateless results: record vendor as "no public funding signal" — frame as classification finding (likely bootstrap or pre-disclosure) rather than fabricating amounts. Don't push for an answer Perplexity can't substantiate.
  • Perplexity citations 404 / URL verification fails: strip silently, log to _internal/url-verification-log.md. Categorical absence of verifiable citations across all vendors flips category-classification toward "bootstrap-dominant" (categorical-zero as positioning signal).
  • Crunchbase / PitchBook citations bot-blocked (403 / CloudFront): these platforms reliably block scrapers — record as "secondary-source verification unavailable" and rely on press release / vendor blog / SEC filing citations only.

What NOT To Do

  • Don't fabricate funding amounts. URL Verification is mandatory.
  • Don't quote Crunchbase if you couldn't actually verify (Crunchbase typically 403-blocks scrapers).
  • Don't extrapolate ARR from headcount without source — most "estimated $X ARR" claims are speculative.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.