| name | thesis-tracker |
| description | Track an investment thesis across time, news flow, and market updates; classify it as strengthened, weakened, unchanged, or falsified. |
| metadata | {"marketbot":{"emoji":"🧠","triggers":["thesis tracker","track thesis","signal journal","view tracking","观点跟踪","逻辑跟踪"," thesis update","strengthened or weakened"],"output":"thesis-tracker-report","risk":"medium","freshness":"event-live","tools":["thesis_tracker","market_news","market_snapshot","intel_search"],"required_tools":["thesis_tracker"],"markets":["a-share","hong-kong","us","global","mixed"],"asset_classes":["equity","etf","commodity","crypto","macro"],"task_type":"thesis-tracking","determinism":"tool-backed","priority":92}} |
Thesis Tracker
Use this skill to persist an investment thesis, revisit it later, and judge
whether new evidence strengthens, weakens, leaves unchanged, or falsifies the
original view.
When to use
- The user wants to track a thesis over days or weeks instead of getting a
one-off report.
- A catalyst, earnings event, or macro shock needs to be judged against a
previously stated view.
market-report, stock-watch, or catalyst-tracker surfaces a view that
should become a durable research object.
Workflow
- Create the thesis with
thesis_tracker(action="create", ...).
- When updating, gather fresh evidence with:
market_news
market_snapshot
intel_search for prior workspace intel
- Update the thesis with
thesis_tracker(action="update", thesisId=..., evidence=..., note=...).
- Use the returned verdict and confidence shift to explain whether the thesis
is:
strengthened
weakened
unchanged
falsified
Output format
# Thesis Tracker: <SYMBOL>
## Current Thesis
- Thesis:
- Status:
- Confidence:
## Latest Update
- Verdict:
- Evidence summary:
- Why the status changed:
## Next Things To Watch
- Catalyst 1
- Risk 1
Rules
- Separate the original thesis from new evidence.
- If the evidence is mixed, prefer
unchanged or weakened over dramatic
reclassification.
- Use
falsified only when the new facts directly break the core premise.
- If evidence is thin, say confidence is low instead of forcing a strong verdict.