Get card
Download
hypothesis-generator@1.0.0.yamlClone
npx -y darkprint clone hypothesis-generator@1.0.0There is no repository and no history behind a card. Download hands you the document as it stands, and Clone fetches the same document by name.
Propose a spread of candidate hypotheses against the agenda, without knowing how they will be graded.
used in 1 blueprint
Specification
120 words · handed to the agentRead the agenda and propose candidate hypotheses for the question it scopes. Emit them on hypotheses as a list, each with the claim itself, the reasoning that makes it worth taking seriously, and what would have to be observed for it to be wrong. Propose a spread rather than variations on one idea: two hypotheses that fail for the same reason are one hypothesis. Do not rank them, do not mark one as most likely, and do not go looking for the conditions they will be graded against, you do not have them and shaping a proposal toward a bar you have guessed at is worse than proposing freely. Do not re-propose a direction the agenda says has been tried.
Interfaces
1 in · 1 outInputs
1| Name | Data type | Required | Description |
|---|---|---|---|
| agenda | plan | required | The scope of this tournament and the directions already ruled out. |
Outputs
1| Name | Data type | Description |
|---|---|---|
| hypotheses | json | Candidate hypotheses, each with its reasoning and its falsifier, in no order. |
Dependencies
1The upstream nodes this card expects to receive from. Whenever a blueprint pins this card, each name is checked against a real edge in that graph. A name without a link is not a published card; it refers to a node inside some graph.
Card values
14 declaredWho the node is. The id is the key the DOT pins.
- id
- hypothesis-generator
- name
- Hypothesis Generator
- type
- agent
- phases
- Implementation
What it does, and the prose the agent is handed when the graph runs.
- action16 words
- Propose a spread of candidate hypotheses against the agenda, without knowing how they will be graded.
- spec120 words
- Read the agenda and propose candidate hypotheses for the question it scopes. Emit them on
hypothesesas a list, each with the claim itself, the reasoning that makes it worth taking seriously, and what would have to be observed for it to be wrong. Propose a spread rather than variations on one idea: two hypotheses that fail for the same reason are one hypothesis. Do not rank them, do not mark one as most likely, and do not go looking for the conditions they will be graded against, you do not have them and shaping a proposal toward a bar you have guessed at is worse than proposing freely. Do not re-propose a direction the agenda says has been tried.in full above - model
- whatever the graph supplies
- agent
- Generator
- skill
- not named
- tools
- none
- mcp
- none
- params
- none
What arrives, what leaves, which nodes it expects to hear from, and what may not.
- inputs
- agenda : plan
- outputs
- hypotheses : json
- dependencies
- research-supervisor
- cannot
- acceptance-criteria
- will_not
- rank or mark a favourite among its own proposals, shape a proposal toward a bar it has guessed at
The keys the static analysis reads. Nothing here instructs the agent.
- risk_markers
- none
- notes
- none
The card's own version, and who wrote it.
- version
- 1.0.0
- author
- autogen
- provenance
- not stated
Version history
1 version published- hypothesis-generator@1.0.0currentsha256:629addb0ebbdd7ba6398729da730a412e60f684b13de30d6291de51e39f9431f
pinned byHypothesis Tournament
autogen/hypothesis-tournament
A digest is a fingerprint (SHA-256) of the card's content, computed without the author and provenance fields. The same card from two people gets the same digest; any edit gets a new one.
First published version, so there is nothing to compare yet. Versions are never edited in place: the next change arrives as a new version, and the differences between the two documents are listed here.
Community notes (0)
No notes yet.
Nobody has posted about this node card yet.
Sign in to post a note.