| name | serenity-skill |
| description | Researches Serenity-style AI and semiconductor supplier opportunities by extracting recent X.com insights, mapping hidden supply-chain bottlenecks, scoring candidates, and producing thesis watchlists. Use when the user mentions Serenity, @aleabitoreddit, unknown bottlenecks, AI/semi suppliers, small-cap supplier stocks, photonics, CPO, HBM, or supplier research. |
Research and structure AI/semiconductor supplier ideas using a Serenity-inspired “unknown bottleneck” method: identify underappreciated suppliers, verify evidence chains, score risk/reward, and produce transparent thesis notes, watchlists, and candidate scorecards.
This skill is for research and education only. It must not present ideas as guaranteed returns, personalized financial advice, or direct buy/sell instructions. It should help users learn a repeatable research process.
<quick_start>
For a fresh request, follow workflows/discover-and-score-suppliers.md.
Default assumptions unless the user says otherwise:
- User wants research and methodology, not copy-trading or direct recommendations.
- Focus on AI compute / semiconductor / photonics / power / HBM / test / packaging bottlenecks.
- Use Serenity’s recent X.com posts as inspiration, then independently verify with filings, company releases, and market data.
- Separate thesis quality from entry quality, especially after viral social-media re-ratings.
</quick_start>
<essential_principles>
- Methodology over copy-trading: Extract how Serenity thinks; do not blindly mirror tickers or infer her current position, cost basis, or exit plan.
- Evidence chain first: Prefer filings, product change notices, customer qualification, capacity ramp, government grants, earnings calls, and official company releases over viral posts.
- Bottleneck purity matters: A candidate should sit on a specific constraint in the AI supply chain, not merely mention AI.
- Thesis quality differs from timing: A strong long-term thesis can still be a poor short-term setup after a large re-rating.
- Risk framing is mandatory: Identify dilution, valuation, hype, customer concentration, weak evidence, and thesis-drift risks.
- No financial-advice certainty: Use language like “research candidate,” “watchlist,” “thesis,” “risk,” and “if confirmed,” not “buy,” “guaranteed,” or “sure winner.”
</essential_principles>
If the user’s request is underspecified, ask only the missing high-impact questions:
1. Research scope: AI/semi broadly, photonics/CPO, HBM/test, power/grid, substrates, inspection, or another theme?
2. Desired output: methodology notes, watchlist, candidate scorecards, or an updated thesis review?
3. Risk posture for scoring: conservative, balanced, aggressive, or high-beta bottleneck hunting?
4. Time horizon: near-term catalyst, 6-12 month thesis, or multi-year architecture cycle?
If the current conversation already provides these answers, do not ask again; proceed.
| User intent | Workflow |
|---|---|
| “Find Serenity-style suppliers”, “research potential supplier stocks”, “build a watchlist”, “map bottlenecks” | `workflows/discover-and-score-suppliers.md` |
| “Update the watchlist”, “re-score these names”, “check if still valid” | `workflows/update-watchlist.md` |
| “Prioritize these candidates”, “make a shortlist”, “which theses are strongest” | `workflows/research-shortlist.md` |
After selecting a workflow, MUST read that workflow file, then MUST read every file listed in its <required_reading> before producing the research output.
<reference_index>
references/serenity-methodology.md — Serenity-style unknown-bottleneck framework, methodology v2, and scoring rubric.references/evidence-standards.md — Source hierarchy, verification rules, and red flags.references/research-risk-and-sizing.md — Public risk framing, thesis sizing concepts, and watchlist status labels.
</reference_index>
<templates_index>
templates/candidate-scorecard.md — Reusable scorecard for each ticker.templates/watchlist-output.md — Final watchlist / shortlist output format.
</templates_index>
<success_criteria>
The skill succeeds when the response includes:
- A clear Serenity-style methodology summary.
- A scored candidate table with bottleneck, evidence, valuation/risk, catalyst, and thesis status.
- Separation between high-conviction research candidates, watchlist candidates, and weak/excluded ideas.
- Explicit risks and “what would invalidate the thesis.”
- A concise next action: verify evidence, wait for a clearer catalyst, read filings, or keep on watchlist.
</success_criteria>
