Get card
Download
acceptance-tester@1.0.0.yamlClone
npx -y darkprint clone acceptance-tester@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.
Run every criterion against the submitted build and split the outcome: the build itself when all of them pass, the raw failure evidence when any of them does not.
used in 1 blueprint
Specification
112 words · handed to the agentYou are handed a set of criteria and a build. Run each criterion against that build in turn, one at a time, and record whether it held. When every one of them holds, emit the build unchanged on approved and emit nothing on evidence. When any of them does not, emit on evidence only what the run itself produced, the assertion that failed, the stack trace underneath it, and the expected value beside the value actually obtained, and nothing else: do not quote the criterion, do not summarise the ones still outstanding, and do not say how many there are in total. A criterion you were unable to evaluate has not held.
Interfaces
2 in · 2 outInputs
2| Name | Data type | Required | Description |
|---|---|---|---|
| criteria | acceptance-criteria | required | The conditions the build is judged against, as the planner wrote them. |
| build | code | required | The submitted change, the builder's first pass, or the debugger's patched version. |
Outputs
2Dependencies
3The 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
17 declaredWho the node is. The id is the key the DOT pins.
- id
- acceptance-tester
- name
- Acceptance Tester
- type
- validation
- phases
- Testing
What it does, and the prose the agent is handed when the graph runs.
- action29 words
- Run every criterion against the submitted build and split the outcome: the build itself when all of them pass, the raw failure evidence when any of them does not.
- spec112 words
- You are handed a set of criteria and a build. Run each criterion against that build in turn, one at a time, and record whether it held. When every one of them holds, emit the build unchanged on
approvedand emit nothing onevidence. When any of them does not, emit onevidenceonly what the run itself produced, the assertion that failed, the stack trace underneath it, and the expected value beside the value actually obtained, and nothing else: do not quote the criterion, do not summarise the ones still outstanding, and do not say how many there are in total. A criterion you were unable to evaluate has not held.in full above - model
- claude-sonnet-5
- agent
- Acceptance tester
- skill
- skills/acceptance-tester.md
- tools
- none
- mcp
- none
- params
- stop_on_first_failure: false
What arrives, what leaves, which nodes it expects to hear from, and what may not.
- inputs
- criteria : acceptance-criteria, build : code
- outputs
- evidence : report, approved : artifact
- dependencies
- spec-planner, code-builder, targeted-debugger
- cannot
- no type is refused
- will_not
- quote a criterion in the evidence it emits
The keys the static analysis reads. Nothing here instructs the agent.
- risk_markers
- none
- notes77 words
- This is the only node that sees both halves, and the shape of
evidenceis what keeps that safe. Doc 2 §5.5: seeing the criteria lets a node write work built to pass them, seeing the evidence of a failure it caused only says what broke, the first is gaming, the second is feedback. Emitting the criterion text alongside the trace would collapse that distinction and hand the debugger the acceptance surface a line at a time.
The card's own version, and who wrote it.
- version
- 1.0.0
- author
- autogen
- provenance
- not stated
Version history
1 version published- acceptance-tester@1.0.0currentsha256:1a7a5e6366d04e276507ccc5d8a2fb64372ebb7e09a014d7283a0ae4756d6881
pinned byStarter Software Factory
autogen/starter-software-factory
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.