rig-competition

A portable rig for studying your competition. Tracks the competitors in your field, surfaces what they ship and what the numbers look like, and flags the moves that pressure your own positioning.

Vertical-agnostic — works for fintech, devtools, B2B SaaS, consumer apps, marketplaces, anything with a definable competitive set and a definable lens of "what matters."

What it does

• Runs a weekly brief on the past 7 days across all tracked comps
• Maintains per-competitor dossiers that accumulate over time (append-only)
• Pulls sourced metrics for one comp or the whole set
• Scans AI/agent features across the field
• Maps the landscape in a single snapshot
• Filters week's moves for threats to your own positioning

Hard rule: no invented numbers, no invented news. Every claim cites a source.

Setup (run this first)

rig use hub:hugo/rig-competition --dir ./my-competition-rig
cd my-competition-rig
claude

Then say setup to the agent. It walks you through:

1. What your business does
2. Your field / vertical
3. Your competitive set (you give names, it researches URLs / IR pages / handles)
4. Your positioning vectors (what a competitor would have to do to threaten you)
5. The reputable trade-press outlets for your field
6. How to bucket the competitive set (2–4 groups)

Output: every file in data/ is filled, cited, and ready. Takes ~15–30 min depending on how many competitors you bring.

After setup, type any of the triggers below.

Skills

Skill	Trigger	Output
setup (run first)	"setup"	Fills data/00–04 from your business description
weekly-brief	"weekly brief"	outputs/weekly/YYYY-MM-DD_brief.md + appends to dossiers
competitor-deep-dive	"deep dive on <comp>"	outputs/competitors/<slug>.md (header refresh + new block)
metrics-pull	"pull metrics for <comp>"	Table in chat + appended to dossier
ai-feature-scan	"AI feature scan past 30 days"	outputs/snapshots/YYYY-MM-DD_ai-feature-scan.md
landscape-map	"show me the landscape"	outputs/snapshots/YYYY-MM-DD_landscape.md
threat-watch	"any threats" (or auto from weekly-brief)	Appended to weekly brief under ## Threats to watch

Context (data/)

File	What's in it
00-tracking-thesis.md	What the rig watches and why; your lens (which signals matter most); cadence rules
01-competitive-set.md	The comps grouped by bucket, with URLs, filings/IR, social, blogs
02-metrics-rubric.md	Per-category metrics that matter; where to source each; what to do when a number isn't public
03-source-rules.md	Citation discipline; your field's allow-list and a universal disallow-list of sources
04-your-positioning-brief.md	Your product's positioning + the threat vectors threat-watch looks for

After setup, edit any of these by hand to add a comp, tighten a source rule, or revise a vector. Skills read these files but don't rewrite them — only setup and you do.

Output

• outputs/weekly/YYYY-MM-DD_brief.md — one per weekly run
• outputs/competitors/<slug>.md — append-only dossier per comp; each block is dated and signed by the skill that wrote it
• outputs/snapshots/YYYY-MM-DD_<type>.md — landscape and ai-feature-scan snapshots

Outputs are gitignored and not packaged.

Customising further

The skills themselves (under skills/) are deliberately generic. If your field needs a different cadence, a different metric matrix, or different threat vectors than what setup produced, edit data/ first — skills read from data/, so most behaviour is data-driven. Touch skills/ only when you want a structurally different output (e.g., monthly briefs instead of weekly).

Version

0.1.0