Investment Thesis Tracker
Keep portfolio theses current, testable, and review-ready.
The skill is limited to maintaining investment theses and drafting outputs; the repository README requires qualified-professional review and disclaims transaction execution, so no clear external side-effect red line is visible. Points are deducted because user confirmation, sensitive financial-data handling, permission boundaries, cross-session storage implementation, rollback, and data-flow disclosure are unspecified; position-sizing and exit actions also lack an explicit human-confirmation gate.
The workflow is internally coherent and defines fields for thesis creation, updates, scorecards, catalysts, and output. Points are deducted for no abnormal-input handling, missing-data policy, date or valuation conventions, storage-failure feedback, or reproducible tests; static review found no skill-specific execution evidence, so the score remains below the static ceiling.
Triggers, audience scenarios, and primary inputs and outputs are reasonably clear for reviewing one or more positions. Points are deducted because non-fit boundaries, source-data requirements, complex-security coverage, trigger priority, and Chinese-language output conventions are not defined. The core function does not depend on a specific overseas service, so no additional mainland-network deduction is warranted.
The documentation is readable and workflow-oriented, while repository-level material identifies the Apache-2.0 license, installation paths, and maintenance checks. Points are deducted because the skill lacks its own version, changelog, named owner, update path, dependency notes, FAQ, troubleshooting guidance, and explicit storage design; limitations are only partially disclosed.
The workflow covers thesis setup, evidence updates, catalyst tracking, and summaries for meetings or risk committees, with a directly usable Markdown or Word output target. Points are deducted because persistence and retrieval are not implemented, data validation and citation format are undefined, human verification remains necessary, and no representative execution evidence is present; the static cap therefore limits the score.
The skill provides auditable fields, steps, and examples, and the repository includes general plugin-validation and secret-scanning workflows. Points are deducted because those CI workflows do not cover the skill's key paths, and there is no committed test suite, third-party execution evidence, or multi-source fact corroboration; only limited static verifiability is supported.
- Do not treat “Increase position / Trim / Exit” as automatic trading instructions; require user confirmation and qualified compliance or investment review.
- The skill does not define structured cross-session storage, source citations, data freshness, or recovery behavior; add these controls before production use.
- Outputs may rely on incomplete or unverified user-provided information and should not be used directly as investment advice or a trading basis.
What it does & when to use it
Investment Thesis Tracker helps maintain theses for portfolio positions and watchlist names. It structures a thesis around the core statement, supporting pillars, invalidating risks, catalysts, valuation, and exit conditions, then logs new developments and their impact. It also maintains a thesis scorecard and catalyst calendar for morning meetings, portfolio reviews, and risk committee presentations. The skill is part of the equity-research vertical plugin in anthropics/financial-services, a repository licensed under Apache-2.0.
For a new thesis, it records the company and ticker, long or short position, a one- to two-sentence thesis statement, three to five key pillars, three to five key risks, upcoming catalysts, target price or valuation, and the stop-loss trigger. For an existing thesis, it asks the user for a new data point or development and records the date, change, affected pillar, action, and updated conviction. It maintains a scorecard with each pillar's original expectation, current status, and trend, plus a catalyst calendar with dates, events, expected impact, and notes. It produces a concise thesis summary in Markdown or Word-document format for meetings, portfolio reviews, or risk committee presentations.
- An equity analyst adds an earnings result, management departure, or competitor move to an existing company thesis.
- A portfolio manager conducts a quarterly review to determine whether each holding's key pillars remain on track and whether conviction is high, medium, or low.
- A research team catalogs upcoming earnings, product launches, and regulatory decisions across its coverage list and assesses how they could confirm or disprove each thesis.
- A risk committee prepares meeting materials with recent updates, thesis scorecards, key risks, and exit triggers.
- An investor managing multiple positions requests a full portfolio thesis review.
Pros & cons
- Covers the full structure from thesis creation through updates, scoring, and catalyst tracking.
- Explicitly requires both confirming and disconfirming evidence and emphasizes that a thesis should be falsifiable.
- Defines outputs for concrete workflows including morning meetings, portfolio reviews, and risk committee presentations.
- Supports both individual thesis updates and a broader review of multiple positions, with at least quarterly review recommended.
- The SKILL.md defines no data source, web research process, or external market-data connector; users must provide new data points.
- It asks for structured storage that can be referenced across sessions but does not specify a storage system or persistence implementation.
- It includes no test suite, automated validation, or detailed valuation methodology.
- Its action field includes increase, trim, and exit options, but the skill does not provide an investment-advice compliance or professional-review workflow.
How to install
The skill is located at plugins/vertical-plugins/equity-research/skills/thesis-tracker/SKILL.md. In Cowork, open Settings → Plugins → Add plugin, paste https://github.com/anthropics/financial-services, and select the equity-research vertical from the marketplace list. In Claude Code, run claude plugin marketplace add anthropics/financial-services, followed by claude plugin install equity-research@claude-for-financial-services. The source does not document a separate installation procedure for copying this individual SKILL.md into another client.
How to use
After installing the equity-research plugin, use a natural-language trigger such as update thesis for ACME, is my thesis still intact, thesis check, add data point to ACME, or review my positions. For a new thesis, provide the company, ticker, position direction, thesis statement, pillars, risks, catalysts, valuation, and stop-loss condition. For an update, provide the new data point or development. The resulting review should include recent updates, the thesis scorecard, the catalyst calendar, and the current conviction level.