Get card
Download
policy-screen@1.0.0.yamlClone
npx -y darkprint clone policy-screen@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.
Put the screened message past the external content policy service and act on what it says.
used in 1 blueprint
Specification
120 words · handed to the agentSend the message to the configured moderation endpoint and read the verdict back. When the service clears it, emit the message on cleared unchanged and set your preferred label to cleared. When the service flags it, emit the categories it returned on flagged and set your preferred label to anything else. When the service is unreachable, times out, or returns something you cannot parse, treat that as flagged rather than as cleared, and say on flagged that the screen did not run: an unavailable check is not a passed check. Send only the message itself, no account identifier and no run metadata. Do not form your own opinion about the content and do not overrule the verdict in either direction.
Interfaces
1 in · 2 outInputs
1| Name | Data type | Required | Description |
|---|---|---|---|
| clean | text | required | The screened message, as the sanitiser passed it on. |
Outputs
2Dependencies
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
16 declaredWho the node is. The id is the key the DOT pins.
- id
- policy-screen
- name
- Policy Screen
- type
- tool
- phases
- none declared
What it does, and the prose the agent is handed when the graph runs.
- action16 words
- Put the screened message past the external content policy service and act on what it says.
- spec120 words
- Send the message to the configured moderation endpoint and read the verdict back. When the service clears it, emit the message on
clearedunchanged and set your preferred label tocleared. When the service flags it, emit the categories it returned onflaggedand set your preferred label to anything else. When the service is unreachable, times out, or returns something you cannot parse, treat that as flagged rather than as cleared, and say onflaggedthat the screen did not run: an unavailable check is not a passed check. Send only the message itself, no account identifier and no run metadata. Do not form your own opinion about the content and do not overrule the verdict in either direction.in full above - model
- whatever the graph supplies
- agent
- Policy screen
- skill
- not named
- tools
- HTTP fetch
- mcp
- moderation
- params
- timeout_ms: 2000, fail_closed: true
What arrives, what leaves, which nodes it expects to hear from, and what may not.
- inputs
- clean : text
- outputs
- cleared : text, flagged : status
- dependencies
- input-sanitizer
- cannot
- no type is refused
- will_not
- treat an unreachable service as a pass, overrule the verdict it is given, send account identifiers or run metadata to the service
The keys the static analysis reads. Nothing here instructs the agent.
- risk_markers
- unvalidated-external-access
- 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- policy-screen@1.0.0currentsha256:36f34ce1a9fd63b1d63b933879550cb969176e951812db540b564f1ca5ffca89
pinned byGuarded Assistant Line
autogen/guarded-assistant-line
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.