CompetLab Agent Skills
What they are
An Agent Skill is a playbook an AI coding agent loads and follows — a SKILL.md file
(following the agentskills.io open standard) that describes a task,
when to use it, and how to do it well. CompetLab Agent Skills package the work of a
competitive-intelligence analyst this way: each skill knows which CompetLab data to pull, how
to combine it with live web research, and how to shape the result into a deliverable you’d
actually hand to a founder, a PMM, or a sales team.
The MCP server gives your agent the raw tools; Agent Skills give it the playbook. On its own, an agent with MCP access can read your dimensions and competitors, but you still have to tell it what a good landscape analysis or battlecard looks like. A skill carries that method with it — so “write a battlecard for Rival Inc” produces a structured, sales-ready card instead of an improvised summary.
The suite
Thirteen skills, in three groups. One orchestrator composes a report from most of the others; five skills work from the data CompetLab already monitors; seven research live signals that lead the monitored dimensions. (These groups describe how each skill works — they aren’t CompetLab’s monitored dimensions. A single skill often draws on several dimensions at once.)
Orchestrator
| Skill | What it produces |
|---|---|
competlab-cmo-report | A complete CMO report — composed by sequencing the dimension dashboards and most of the other skills into one main report, a dozen per-dimension docs, and per-competitor deep dives. |
Lagging-indicator skills — work from your CompetLab monitoring data (what’s already measured):
| Skill | What it produces |
|---|---|
competlab-ai-visibility | How ChatGPT, Claude, and Gemini mention and recommend your brand versus competitors, from your AI Visibility Score data. |
competlab-weekly-briefing | An on-demand delta briefing — what changed, what it means, what to do — that you run whenever you want a pulse. |
competlab-competitor-dive | A deep single-competitor dossier: tech, pricing, positioning, content, AI visibility, sentiment, SWOT, and recommended responses. |
competlab-battlecard | A sales-ready battlecard for a competitor — quick-reference for calls, demos, and objection handling. |
competlab-landscape | A full competitive landscape across all five monitored dimensions, with market dynamics and cross-dimensional patterns. |
Leading-indicator skills — research live public signals (what competitors are about to do):
| Skill | What it produces |
|---|---|
competlab-status-watch | Reliability read from each competitor’s public status page and incident history. |
competlab-funding-watch | Recent funding, ownership changes, ARR estimates, M&A, and executive transitions. |
competlab-ai-ecosystem | External developer-ecosystem signals — GitHub, npm/PyPI volumes, community MCP servers, marketplace presence. |
competlab-hiring-signals | Hiring and go-to-market motion from public ATS listings, plus executive-transition research. |
competlab-agent-adoption | Agent Adoption posture, wrapping CompetLab’s scan and adding protocol-level verification of any MCP endpoints it finds. |
competlab-product-watch | A snapshot of a competitor’s product surface beyond the homepage — changelogs, releases, API-doc versions. |
competlab-customer-voice-snapshot | A snapshot of review-platform presence (G2, Capterra, Trustpilot) and recent complaint themes. |
Every skill, with what it needs and when to reach for it, is in the skill reference.
How they work
A skill reads your data through the CompetLab MCP server — it calls the mcp__competlab__*
tools, the same ones documented in the MCP tool reference. The
leading-indicator skills go further, shelling out to public sources (GitHub, npm, ATS APIs,
status-page backends) and using web research to gather signals CompetLab doesn’t monitor
directly, then verifying every URL before it lands in a deliverable. The output is always
plain Markdown — a briefing, a dossier, a card — that you can read, edit, and share.
Because the skills run on the MCP server, they see exactly what your API key can see and nothing more.
The MCP connection is a hard dependency — without it configured, a skill either degrades to web-research-only or tells you it can’t run. Set it up with the Connect guide.
What you need
- A CompetLab account with at least one project and its competitors being monitored.
- The CompetLab MCP server configured with your API key — this is how the skills reach your data. See the Connect guide to set it up.
- A compatible agent. The skills are built for Claude Code (the plugin marketplace and the
mcp__competlab__*tool names are Claude Code conventions). Because they’re standardSKILL.mdfiles, other agents can use them too — you may just need to adjust the MCP tool names for your client.
Authentication lives in the MCP configuration, not the skills — there’s no separate skill-level API key. The Setup guide walks the whole thing.
What they don’t do
A few honest limits:
- They read your account; they don’t change it. Skills pull your monitoring data and run the public scans; nothing a skill does writes to or changes your CompetLab projects, alerts, or settings.
- Coverage varies by skill, by design. Some are explicitly snapshot-only —
product-watchandcustomer-voice-snapshotdon’t produce velocity or alerts, because the underlying public sources (changelogs, review platforms) are partial or bot-blocked. Each skill states its own limits. - The research skills depend on public sources. The leading-indicator skills are only as complete as what a competitor exposes publicly — a missing status page or a private ATS means less signal (and sometimes that absence is itself the finding).
- The weekly-briefing skill is user-run, not a product cadence. It’s an on-demand delta you trigger whenever you want; it’s distinct from CompetLab’s monthly Strategic Briefing.
Next steps
- Setup → — install the plugin, connect the MCP server, and run your first skill.
- Skill reference → — all 13 skills with what each needs and when to use it.
FAQ
What are CompetLab Agent Skills?
They're a suite of 13 prebuilt skills — the open-source competlab-ci-skills plugin — that turn your CompetLab data into finished competitive-intelligence deliverables. An Agent Skill is a SKILL.md playbook (following the agentskills.io open standard) that an AI coding agent loads and follows. Each CompetLab skill knows which data to pull from the MCP server, how to combine it with live web research, and how to shape the result into something you'd hand to a founder, a PMM, or a sales team — a landscape analysis, a competitor dossier, a battlecard, an AI-visibility report, an on-demand briefing, or a full CMO report.
How are Agent Skills different from the MCP server?
The MCP server gives your agent the raw tools; the skills give it the playbook. With MCP access alone, an agent can read your dimensions and competitors, but you still have to tell it what a good deliverable looks like. A skill carries that method with it — so a plain-language request like "write a battlecard for Rival Inc" produces a structured, sales-ready card instead of an improvised summary. The skills run on top of the MCP server: they call the same mcp__competlab__* tools, so the MCP connection is a prerequisite for using them.
What's in the suite?
Thirteen skills in three groups. One orchestrator (competlab-cmo-report) composes most of the suite into a live CMO report. Five lagging-indicator skills work from your CompetLab monitoring data: ai-visibility, weekly-briefing, competitor-dive, battlecard, and landscape. Seven leading-indicator skills research live public signals that lead the monitored dimensions: status-watch, funding-watch, ai-ecosystem, hiring-signals, agent-adoption, product-watch, and customer-voice-snapshot. The skill reference documents each one.
How do I install them?
Three ways. In Claude Code, add the plugin marketplace with /plugin marketplace add competlab/competlab-ci-skills and install with /plugin install competlab-ci-skills@competlab-ci-skills. With the skills CLI, run npx skills add competlab/competlab-ci-skills --all (that installs the skill folders but not the shared companion docs the CMO-report orchestrator reads). Or clone the repo and copy the folder contents with cp -r competlab-ci-skills/skills/. .claude/skills/ — the trailing /. also brings those companion docs. Before any of them do useful work, you also need the CompetLab MCP server configured with your API key — that's how the skills reach your data. The Setup guide covers all three paths plus the MCP prerequisite.
How do the skills authenticate to CompetLab?
Through the MCP server, not the skills themselves. The skills call the mcp__competlab__* tools, and authentication is handled entirely by your MCP server configuration — the API key you set up there. There's no separate skill-level API key or environment variable. A skill sees exactly what your key can see and nothing more. (One optional exception: a couple of the research skills accept a GH_TOKEN to lift GitHub's unauthenticated rate limit when scanning developer-ecosystem signals.)
Do the skills change anything in my CompetLab account?
No. The skills read your monitoring data and run the public scan tools; nothing a skill does writes to or changes your projects, competitors, alerts, or settings. Their output is plain Markdown deliverables — briefings, dossiers, cards — that you read, edit, and share. The only "writes" anywhere in the CompetLab surface are the free scan tools that start a scan of a URL you provide, which some skills use.
Which agents can run them?
They're built for Claude Code — the plugin marketplace install and the mcp__competlab__* tool naming are Claude Code conventions. Because they're standard SKILL.md files following an open standard, other coding agents can use them as well, but you may need to adjust the MCP tool names to match how your client namespaces them. Whatever the agent, the CompetLab MCP server has to be configured for the skills to reach your data.
Is the weekly-briefing skill the same as the monthly Strategic Briefing?
No — they're different things. The weekly-briefing skill is an on-demand delta you run yourself whenever you want a pulse on what changed; many teams make it a weekly habit, but that's a user choice, not a product cadence. CompetLab's Strategic Briefing is the synthesized cross-dimensional read the platform produces on a monthly cadence — you pull it through the get_briefing MCP tool. The cmo-report skill fetches that briefing and composes a broader CMO report around it — the report is a superset that includes the briefing, and the skill fetches the Strategic Briefing rather than producing it. The weekly-briefing skill is lighter and delta-focused; the Strategic Briefing is broader and periodic.