- Home
- Skills
- Trading & Investing
- Investment Thesis Sanity Gate
Investment Thesis Sanity Gate
Turn a messy investment idea into an evidence-first thesis audit with weak points, disconfirming data, and next research actions.
Free
Works with the AI Agents you already use
Investment Thesis Sanity Gate
Example session with this skill installed
Use investment-thesis-sanity-gate on this idea. Asset: Northstar Pumps, a fictional industrial pump manufacturer. Thesis: EBITDA can grow faster than revenue over the next 18 months because supply-chain costs are normalizing, aftermarket mix is rising, and leverage should fall enough for a valuation re-rating. Time horizon: 12-18 months. Evidence: FY2025 revenue +7% and adjusted EBITDA margin 12.4% vs 14.8% pre-disruption, dated 2026-03-12. Q1 2026 orders +9%, aftermarket revenue +16% to 31% of sales, and management comment that supplier expedite fees should normalize by Q4 2026, dated 2026-05-08. User peer table dated 2026-05-10 shows pump peers at 9-11x NTM EBITDA while Northstar is estimated at 7.8x. Current concern: The company has 42% revenue exposure to chemical and energy capex, plus a 2027 debt maturity. Am I underwriting a recovery that may already be priced in?
- Read your context and instructions
- Compiled the investment thesis sanity
Investment Thesis Sanity Gate
Thesis Snapshot
Northstar has a plausible margin-recovery thesis, but the evidence is not yet decision-ready. The upside depends on EBITDA margin recovery, aftermarket mix durability, and a peer-multiple re-rating. The main downside risks are cyclical capex exposure and refinancing risk.
Claim Audit
| Claim Type | Status | Notes |
|---|---|---|
| Business | Supported | Aftermarket revenue grew 16% and reached 31% of sales, but durability through a downturn is not proven. |
| Financial | Weak | The 12.4% to 15% margin bridge needs cost, price, volume, and mix math. |
| Valuation | Weak | The 7.8x to 9x re-rating needs peer adjustments for leverage, growth, margin, and cyclicality. |
| Risk | Supported | Chemical/energy capex exposure and the 2027 debt maturity are material risks. |
Disconfirming Evidence to Find
- Backlog conversion and cancellation rates by end market.
- Gross margin bridge separating supplier expedite fees from mix and pricing.
- Refinancing terms, covenant headroom, and interest expense sensitivity.
Decision Gate
Verdict: Proceed to deeper research. This is a research green light, not a buy/sell/hold recommendation.
Connects securely to your tools. The creator never sees your data.
About this skill
What it does
The Investment Thesis Sanity Gate acts as a rigorous quality-control filter for equity research and financial analysis. It transforms draft investment notes, watchlist ideas, or pre-IC (Investment Committee) memos into a structured audit that tests the strength of your conviction. By decomposing your thesis into verifiable claims, the skill separates objective evidence from speculative assumptions and logical inferences.
Why use this skill
Even for experienced analysts, it is easy to fall into the trap of "story-based investing" or ignoring disconfirming data. This skill is better than manual prompting because it follows a systematic Phase-Gate methodology: checking input completeness, auditing evidence quality by date, and running specific stress tests like 'priced-in' expectations and 'what would make me change my mind.' It moves beyond simple summarization to proactively identify research gaps.
Supported Outputs
- Thesis Snapshot: A high-level mapping of drivers, risks, and evidence timelines.
- Claim Audit: A status-coded breakdown (Supported/Weak/Contradicted) of business, market, financial, and catalyst claims.
- Evidence Map: Categorization of inputs into Evidence, Assumptions, and Inferences with confidence levels.
- Decision Gate: A final research verdict (e.g., Proceed, Watchlist, or Reject) with specific next-step research actions.
This skill provides decision support, not personalized financial advice. It does not issue buy/sell/hold instructions, target prices, or guaranteed return claims. Output quality depends on the user's supplied thesis, evidence, dates, valuation assumptions, and risk notes. It does not fetch live market data unless the user's agent separately has browsing or data tools enabled.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
This skill has been used 15 times.
Trust & safety
Security scan
Security scan passed 4 months ago
- Free with an Agensi account
Permissions
No special permissions declared or detected