Skip to content
DarkPrint
Aautogencourse-corrector1.0.0

Course Corrector

Get card

Clone

npx -y darkprint clone course-corrector@1.0.0

There 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.

↓ 0 downloads

Replan around what drifted, or say plainly that the run is blocked and stop.

used in 1 blueprint

Specification

147 words · handed to the agent

You are handed a drift signal. Work out what would put the run back on course and emit the revised steps on revised_steps, in the same shape the planner uses, so the executor takes them the same way. Change only what the drift names: a correction that rewrites the whole plan is a new plan, and the run stops being the one the owner asked for. When the drift names something you cannot correct from here, a missing permission, a dependency outside the run, an objective that turns out to be unreachable as stated, do not invent a workaround: emit on stuck what is blocking the run and set your preferred label to blocked, so a person sees it. Otherwise set your preferred label to anything else. You do not have the conditions the run is judged against and must not reconstruct them from the drift signal.

Interfaces

1 in · 2 out

Inputs

1
Inputs declared by this node card
NameData typeRequiredDescription
driftsignal requiredWhat is moving away from the objective, as the monitor described it.

Outputs

2
Outputs declared by this node card
NameData typeDescription
revised_stepsplanThe corrected steps, changed only where the drift named something.
stuckstatusWhat is blocking the run, when correcting it from here is not possible.

Dependencies

1

The 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 declared

Who the node is. The id is the key the DOT pins.

id
course-corrector
name
Course Corrector
type
agent
phases
Debugging
Behaviourcard spec →

What it does, and the prose the agent is handed when the graph runs.

action14 words
Replan around what drifted, or say plainly that the run is blocked and stop.
spec147 words
You are handed a drift signal. Work out what would put the run back on course and emit the revised steps on revised_steps, in the same shape the planner uses, so the executor takes them the same way. Change only what the drift names: a correction that rewrites the whole plan is a new plan, and the run stops being the one the owner asked for. When the drift names something you cannot correct from here, a missing permission, a dependency outside the run, an objective that turns out to be unreachable as stated, do not invent a workaround: emit on stuck what is blocking the run and set your preferred label to blocked, so a person sees it. Otherwise set your preferred label to anything else. You do not have the conditions the run is judged against and must not reconstruct them from the drift signal.in full above
model
whatever the graph supplies
agent
Corrector
skill
not named
tools
none
mcp
none
params
max_retries: 2
Interfacescard spec →

What arrives, what leaves, which nodes it expects to hear from, and what may not.

inputs
drift : signal
outputs
revised_steps : plan, stuck : status
dependencies
progress-monitor
cannot
acceptance-criteria
will_not
rewrite the whole plan to answer one drift signal, invent a workaround for something outside the run
Evaluation metadatacard spec →

The keys the static analysis reads. Nothing here instructs the agent.

risk_markers
none
notes21 words
max_retries bounds the correct and re-execute loop. Without it this node and the executor sit in a cycle with no cap.
Service fieldscard spec →

The card's own version, and who wrote it.

version
1.0.0
author
autogen
provenance
not stated

Definitions for every card field

Version history

1 version published
  1. course-corrector@1.0.0currentsha256:7334215dca725883a30ae6c9f00cf7a760d8cbf97a6d6fd3bdc0986eff8efd9b

    pinned byObjective Trackerautogen/objective-tracker

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.