Know what's been built. Find your gap.
An AI skill that turns your coding agent into a Stellar ecosystem analyst — for builders entering hackathons, applying for SCF grants, or shipping independently.
Installable into any agent that loads SKILL.md files: Claude Code, Codex, Cursor, OpenClaw, Amp, Antigravity, Cline, and 50+ others via the skills CLI.
# Claude Code (default)
npx skills add Stellar-Light/stellar-scout
# Codex
npx skills add Stellar-Light/stellar-scout -a codex
# OpenClaw
npx skills add Stellar-Light/stellar-scout -a openclawOr paste the raw SKILL.md into your agent's skill directory.
Pre-hackathon idea vetting
"I want to build a Stellar wallet for college students with instant USDC off-ramps in Latin America. Should I?"
Scout pulls existing wallet projects, checks SCF funding patterns, surfaces past hackathon winners in adjacent space, classifies the gap (full / partial / false), and tells you honestly whether to differentiate or pick a different lane.
SCF grant prep
"Is there an open Stellar RFP that matches my idea — a real-time price API for Soroban tokens?"
Scout queries the live RFP feed, finds the matching brief, surfaces author + tech requirements + funding terms, and warns when an RFP is closed for the current quarter.
Security-conscious design
"I'm designing a Soroban lending market. What audit findings should I worry about?"
Scout searches a corpus of every published Soroban audit (Certora, OtterSec, Halborn, OpenZeppelin, Code4rena, Veridise, Cantina, Runtime Verification, etc.) and surfaces real findings with auditor + severity + protocol inline — "per the Certora audit of Blend Protocol V2 (HIGH severity), the oracle can be manipulated when …"
Ecosystem questions
"How has Stellar's developer count changed since 2019?"
Scout pulls from the Electric Capital Developer Reports (2019–2023) for hard macro data, not vibes.
Scout teaches your AI agent how to query 14 public read-only APIs on stellarlight.xyz — no auth, edge-cached, rate-limited.
| Endpoint | Purpose |
|---|---|
/api/research |
Vector search over a ~4,500-chunk research corpus (Voyage AI voyage-3 + Atlas Vector Search). Sources: SDF blog, SCF Handbook, SEPs, dev docs, foundational papers, lumenloop community playbooks, Soroban audits, Electric Capital Developer Reports. |
/api/hackathons & /api/hackathons/{slug} |
Curated hackathons + live DoraHacks feed. Empty-state surfaces fallback channels (BuildOnStellar / stellarlight / DoraHacks org). |
/api/hackathons/compare |
Side-by-side compare of 2–5 hackathons with delta notes. |
/api/analyze |
Cross-event analytics: prize totals, dev funnel, top categories, SCF distribution. |
/api/clusters |
Topic clusters across the projects directory with crowdedness scores (log-scaled, 1–10). |
/api/projects/search |
Prior-art + competitor lookup with tiered match-mode (strict → loose → majority). |
/api/rfps |
Open + closed RFPs with quarter awareness. |
/api/builders |
Builder directory (sourced from Stellar Passport). |
/api/skills & /api/skills/{name} |
Catalog of skills.stellar.org's 7 official skills. |
/api/leaderboard |
Ecosystem dev activity, peer L1 comparison. |
/api/feedback |
Submit feedback when the skill gets something wrong. Closes the loop. |
/api/status |
Health check + endpoint enumeration. |
This skill covers strategy: what's already been built, what should you build, who should you talk to.
For execution (writing Rust on Soroban, building a dApp frontend, integrating SEP standards), install the official skills.stellar.org skills alongside this one. The skills compose:
stellar-scout← what to build, what's been built, where's the gap, what's open RFPskills.stellar.org/soroban← how to write the contractskills.stellar.org/dapp← how to wire the frontendskills.stellar.org/anchors← how to integrate with Stellar anchors
Scout grounds answers in primary sources, with inline citations:
- Stellar Development Foundation — blog posts, SCF Handbook, developers.stellar.org
- SEPs — all 57 protocol standards from stellar/stellar-protocol
- Foundational papers — Mazières et al. on the Stellar Consensus Protocol
- lumenloop community — SCF playbooks + ecosystem research
- sorobansecurity.com — every published Soroban audit (13 firms, 56 reports)
- Electric Capital Developer Reports — annual + geographic ecosystem analyses
Run periodically to pull the latest version:
npx skills update stellar-scoutThe corpus auto-refreshes daily at 06:00 UTC. New audit reports, RFPs, and ecosystem posts appear in /api/research within a day of publication.
When the skill gets something wrong, your agent can call POST /api/feedback directly. Or:
- Open an issue in this repo
- Suggest a missing source via stellarlight.xyz/submit
This SKILL.md mirrors stellarlight/public/skills/stellar-scout.md. The skill content lives next to the APIs it documents.
MIT. Built as a public good for Stellar builders by stellarlight.xyz.