@synap-core/workspace-templates 0.12.0 → 0.23.0

package/skills/stellar-scf-grants-advisor/ASSOCIATE-CHEATSHEET.md

@@ -0,0 +1,133 @@
+# Associate cheatsheet (1 page)
+
+Print this. Keep it next to you during reviews and client calls.
+
+## The 7-step review process
+
+1. Read the client's submitted content (every field).
+2. Web-fetch website, deck, founder LinkedIns, GitHub.
+3. Score against 14 signals (1 to 10 each).
+4. Identify rejection patterns + cite past examples.
+5. Surface winning patterns (don't erase strengths in rewrites).
+6. Compare to 10-15 corpus comparables.
+7. Generate Excel + chat summary with probability range.
+
+## The Excel deliverable (4 sheets, always)
+
+| Sheet | Content |
+|---|---|
+| 1. Field-by-Field Review | Q-numbers, Status, Review (FR or EN), Rewrite (EN paste-ready) |
+| 2. Pre-submission Checklist | Priority, Item, Field, Effort, Why, Status |
+| 3. Corpus Comparables | 10-15 past dossiers side-by-side with current |
+| 4. Integration List Audit | Each Stellar item: MENTIONED / TO ADD / OPTIONAL / NOT RELEVANT |
+
+## Banner row formula
+
+> "[Project] [Phase] Review v[N]: estimated XX to YY pct probability now, AA to BB pct after fixes. Budget anchor $[X],000 [Track], anchored on [Named Past Winner] ($YYK SCF #NN)."
+
+## Vocabulary translation (NEVER use left column with clients)
+
+| Internal | Client-friendly |
+|---|---|
+| kill signature | common rejection pattern, recurring red flag |
+| corpus | our database of 130+ rejected and 79+ awarded dossiers |
+| doxxing depth | whether founders are publicly identifiable |
+| bolt-on | port from another chain without a Stellar-specific reason |
+| TAM-grandiose | vague total addressable market claims |
+| sub-max | below the $150K maximum |
+| heightened scrutiny | extra scrutiny applied at maximum-ask submissions |
+
+## The 14 signals (memorize order)
+
+1. Team identity (doxxing depth)
+2. Stellar-native traction
+3. Budget realism vs scope
+4. Scope clarity (named contracts + Integration items)
+5. Differentiation from previously-funded teams
+6. Track choice fit (Integration vs Open vs RFP)
+7. Go-to-market specificity (named partners + customers)
+8. Milestone precision (verifiable evidence + quantified targets)
+9. Traction evidence (live URLs + LOIs + revenue)
+10. Regulatory framing (named framework + license partner)
+11. Open-source posture (MIT/Apache 2.0)
+12. Brand non-collision
+13. SDF roadmap alignment (x402, Protocol 25 X-Ray, EURC, MiCA, agentic payments)
+14. Comeback signal (only for reapplying teams)
+
+## The 12 rejection patterns (memorize by name)
+
+1. Unidentifiable-founder pattern
+2. Maximum-budget-without-scope pattern
+3. Generic-name pattern
+4. Brand-collision pattern
+5. Multi-chain-bolt-on pattern
+6. SDF-roadmap-overlap pattern
+7. Already-funded-competitor pattern
+8. Regulated-product-without-regulatory-plan pattern
+9. Vague-TAM pattern
+10. Earn-via-low-effort-activity pattern
+11. Incomplete-dossier pattern (Information Collection status)
+12. Cluster-of-issues rejection (Panel Review Failed)
+
+## Budget anchors (memorize)
+
+| Anchor | Profile | Budget range |
+|---|---|---|
+| Bando (SCF #42 won) | RWA single-pilot, doxxed, regulated partner named | $75K-$90K |
+| TheXBank (SCF #40 won) | Africa banking with country pilot | $80K-$100K |
+| The Signal (SCF #42 won) | B2B marketplace, 2 doxxed, 1 contract | $115K-$125K |
+| Rebond (target SCF #44) | RWA bonds, 2 doxxed, 4 LOIs, regulatory framework | $125K-$135K |
+| For Yield (target SCF #44) | Regulated DeFi yield, 11-team, AMF filed | $135K-$145K |
+
+Tranche structure (always 3 equivalent tranches):
+- $90K = 3 x $30K
+- $108K = 3 x $36K
+- $121.2K = 3 x $40.4K
+- $132K = 3 x $44K
+- $144K = 3 x $48K
+
+## Pre-screen blockers (NEVER ship a Build phase dossier without these)
+
+- Q5 Code URL: public GitHub with README + MIT + scaffold (ideally testnet contract deployed)
+- Q6 Video URL: YouTube/Vimeo, under 3 min, 5-segment script, live demo on Stellar Expert
+- Q10 Technical Architecture: public Notion or GitHub markdown with C4 L1 + L2 + flows
+- Q8 Thumbnail: 1200x630 image uploaded
+
+## SCF Integration List essentials
+
+- **Wallets**: Stellar Wallets Kit (default), Privy (mainstream), DFNS (institutional), Freighter Connect
+- **Off-ramp**: Anchor Platform (EU), Bridge (Stripe-acquired), MoneyGram (emerging markets), Mercuryo
+- **DeFi**: DeFindex (yield), Blend v2 (lending), Aquarius (DEX), Soroswap (aggregator)
+- **Cross-chain**: Allbridge, Near Intents, Axelar
+- **Payments**: Stellar Disbursement Platform (SDP)
+
+Items NOT on the list (Fireblocks, Elliptic, Certora, Wirex) = supporting infra, not Integration Track items.
+
+## Em-dash rule
+
+Never use the long-dash punctuation character (the wide horizontal line, Unicode U+2014) in any client deliverable. Use commas, colons, parentheses, periods, or single hyphens instead. The long-dash reads as machine-generated.
+
+## QA before sending any deliverable
+
+1. Zero em-dashes anywhere in the file.
+2. Zero use of "kill signature" anywhere in the file.
+3. Every flagged problem cites at least one named past dossier.
+4. Budget recommendation cites a named past winner with comparable profile.
+5. Every paste-ready rewrite (Column D) is in English, not French.
+6. Banner row contains specific numbers (not "TBD").
+7. Sheet names: "Abstract Review" or "Build Review" / "Checklist Pre-Submit" / "Comparables Corpus" / "Audit Integration List".
+
+## The chat summary after every delivery
+
+> "Dossier livré, fichier v[N] sauvegardé.
+> Verdict: estimated XX-YY pct probability now, AA-BB pct after fixes.
+> Top 5 fixes (in priority order): [list].
+> Recommended budget: $[X],000 [Track], 3 tranches of $[X/3].
+> [Open the file](computer://path/to/file.xlsx)"
+
+## Quick prompts for Claude with the skill
+
+- Review: "Review this SCF Abstract: [URL]. Round SCF #[N]. Give me the 4-sheet Excel deliverable."
+- Budget only: "Recommend budget for [category] dossier with [team] and [traction]."
+- Triage: "Top 3 rejection patterns this dossier triggers. Cite past dossiers."
+- Comeback: "Lost SCF #[N] at $[X]. Profile: [details]. What changed? What to fix?