problem-prospector

Mine evidence-backed problem statements from government audits, grievances, and procurement records.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/vinodkrishna221/Q-Trace --skill problem-prospector-vinodkrishna221
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: problem-prospector
Source: https://github.com/vinodkrishna221/Q-Trace/tree/main/.agents/skills/problem-prospector
Command: npx skills add https://github.com/vinodkrishna221/Q-Trace --skill problem-prospector-vinodkrishna221

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Hackathon teams in self-proposed tracks (like SIH Open Innovation) often brainstorm generic problem statements that look identical to every other team's. This Skill replaces brainstorming with evidence mining: it finds problems that institutions have already admitted in official documents, with numbers, links, and named failed attempts. ## Core Features & Use Cases - Five parallel mining pipelines: admission mining (CAG audits, parliamentary questions), grievance mining (CPGRAMS, state portals), procurement mining (tenders, GeM), graveyard mining (failed pilots and startups), and friction mining (recurrence, court orders). - Multi-gate validation: triangulation across 3+ independent streams and 2+ years, an anti-generic chair test, whitespace checks, and a software-native wedge screen ensuring the problem is solvable by code in a 36-hour build. - PQ scoring and dossiers: ranks 8-12 candidates on severity, recurrence, whitespace, solvability, demoability, and deployability, then produces top-3 full dossiers plus an SIH-format problem statement. - Use Case: A team entering SIH's Smart Education theme runs the prospector to translate the theme into concrete sectors, mines CAG reports and state grievance portals, and emerges with three ranked, evidence-backed problem statements instead of a generic "lack of awareness" pitch. ## Quick Start Run the problem prospector in industry mode for agritech and return the top three evidence-backed problem statements with full dossiers.

Frequently Asked Questions about problem-prospector

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find a real problem statement for SIH Open Innovation?

Mine institutional documents instead of brainstorming: CAG audit reports, parliamentary questions, CPGRAMS grievance statistics, and government tenders all contain admitted problems with numbers. A valid candidate needs at least three independent evidence streams across two or more years.

What sources does problem mining use for evidence?

Five pipelines run in parallel: admission mining (CAG reports, scheme evaluations, budget utilization gaps), grievance mining (CPGRAMS, state portals), procurement mining (GeM tenders, grand-challenge archives), graveyard mining (failed pilots and startups), and friction mining (court orders, seasonal recurrence).

How do I avoid generic hackathon problem statements?

Apply the chair test: kill any framing writable without leaving a chair, such as "lack of awareness" or "no unified platform." A survivor must name a specific place, a sourced number, a failed prior attempt with its reason, and the institution that owns the pain.

Can a hardware-heavy theme work for a software-only hackathon team?

Only after reframing. Gate 0 screens cells for software-native value creation before mining begins; "build drones" is rejected while "compliance workflows for organizations that already own drones" survives. Structurally physical themes with no software reframe are flagged before budget is spent.

What makes a mined problem statement score well?

Candidates are PQ-scored 1-5 on severity, recurrence, whitespace, solvability, demoability, and deployability. The weighted sum ranks 8-12 candidates, and the top three receive full dossiers with evidence files, chain-of-loss, and candidate wedges.

When should a well-evidenced problem still be rejected?

Reject it when no software-only wedge exists: if the core value is not produced by code, inputs require custom hardware, the demo depends on physical action, or the build plan involves fabrication. Log it as evidence-strong but software-infeasible to avoid rediscovery.