Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/inscico/i-framework/competenpx skills add InSciCo/i-framework --skill competegit clone --depth 1 https://github.com/InSciCo/i-frameworkWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00068 | $0.01069 |
| Opus 5 | $0.00034 | $0.00535 |
| Sonnet 5 | $0.00014 | $0.00214 |
| Haiku 4.5 | $0.00007 | $0.00107 |
Grade A, and why
compete scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/compete — competitor research & product comparison matrix
You replace the founder's assertion of differentiation with sourced evidence. You research real competitors on the live web and render a comparison matrix that grounds the UVP and the innovation-vs-implementation classification.
Live web research carries hallucination and staleness risk. Two non-negotiables: (1) every factual cell traces to a source URL; (2) anything you can't source is marked ? and listed for the founder to verify. Never present an unsourced guess as fact.
Procedure
1. Load context
Read Product/intent.md (the problem, target user, and the status quo / DIY alternative), Product/ideation.md (the bet), and Product/innovation.md (the UVP elements — UVP-1, UVP-2… become the must-have rows of the matrix). If innovation.md is missing, ask the founder for the differentiators before proceeding.
2. (Optional) seed
Ask if the founder already knows key rivals. Accept a seed list — but still search for the ones they don't know about. Unknown competitors are often the most useful finding.
3. Discover competitors (web search)
Search the live web from several angles: the problem/category, "alternatives to ", " tools/software", " vs", review roundups. Classify what you find into direct, indirect / adjacent, and the status-quo / DIY alternative (pulled from intent.md — often a spreadsheet, a manual process, or "nothing"). Shortlist the 3–5 most relevant rivals so the matrix stays readable.
4. Research each rival (fan-out)
For thoroughness, spawn one subagent per shortlisted rival via the Task tool. Each gathers: positioning, key capabilities, pricing, target segment, strengths, weaknesses — and returns source URLs. Use WebFetch to read product/pricing pages directly rather than trusting search snippets.
5. Verify & source
Every factual claim cites a source. For each capability you can confirm from a source, mark the cell ✓ / ~ / ✗. For anything you cannot source, mark it ? and add it to the verification list. Marketing pages overstate — prefer docs, pricing pages, and third-party reviews over a vendor's own superlatives.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 48 lines · 68 tokens per session scan A cb4bfabe492b
compete is a skill published in the GitHub repository InSciCo/i-framework (4 stars, last pushed 1mo ago), licensed MIT. It adds 68 tokens to every session and 1,069 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
resume-customizer
Director-level / senior PM resume + cover letter toolkit. Generates ATS-optimized documents using verified metrics and company-type strategies. For interview prep, STAR stories, outreach templates, and salary negotiation, see the references/ folder. Personalize references/YOURPROFILE.md first.
patterns
Analyze the application tracker for what converts and what wastes time. Surfaces score thresholds, winning archetypes, and the blockers driving rejections, then proposes targeting changes. Argument: none.
apply
Paste a JD to get full analysis + tailored resume + cover letter. Argument: [paste JD text here].
followup
Follow-up cadence tracker. Flags overdue and urgent applications, then drafts the second-touch email or LinkedIn message in your voice. Argument: none, or a company name to follow up on one.
track-application
Log a new application or check pipeline status. Argument: [company] [role] [status: applied/screening/interview/offer/rejected].
interview
Full interview prep for a specific company. Argument: [company name] [optional: role title].