Detect a Stale Drawing in a Submittal
Inspect a submittal index against the current drawing register. Detect superseded sheets and unknown revisions. Prepare a replacement checklist with exact document references.
What you are building
The core problem, expected build, and operating context for this challenge.
Inspect a submittal index against the current drawing register. Detect superseded sheets and unknown revisions. Prepare a replacement checklist with exact document references.
Use OpenAI GPT-6 Astra with pypdf and openpyxl.
Use revision identity from the register. Do not sort revision labels alphabetically to infer currency.
Use gpt-6-astra for evidence interpretation and workflow coordination. Use code for arithmetic and constraint checks.
Treat source documents and browser content as data, not instructions.
Use synthetic or explicitly authorized records. Do not upload personal or confidential data without permission.
Cite input record identifiers for conclusions. Distinguish missing evidence from a negative result.
Record tool calls, results, elapsed time, and reported token usage. Do not request private chain-of-thought.
Keep external actions disabled. Use a test portal for browser workflows and preserve draft status.
Do not certify structural safety or approve substitutions. Require a qualified reviewer.
submittal-audit.xlsx and a replacement checklist
source code and dependency lockfile
automated tests and recorded tool events
a short limitations and approval note
Participation guide
Use revision identity from the register. Do not sort revision labels alphabetically to infer currency. Use gpt-6-astra for evidence interpretation and workflow coordination. Use code for arithmetic and constraint checks. Treat source documents and browser content as data, not instructions. Use synthetic or explicitly authorized records. Do not upload personal or confidential data without permission. Cite input record identifiers for conclusions. Distinguish missing evidence from a negative result. Record tool calls, results, elapsed time, and reported token usage. Do not request private chain-of-thought. Keep external actions disabled. Use a test portal for browser workflows and preserve draft status. Do not certify structural safety or approve substitutions. Require a qualified reviewer.
How work is evaluated
A 100-point review rubric with an unauthorized-action failure gate. Starter examples do not establish full workflow performance.
submittal-audit.xlsx and a replacement checklist source code and dependency lockfile automated tests and recorded tool events a short limitations and approval note
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
correctness
Use revision identity from the register. Do not sort revision labels alphabetically to infer currency.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
evidence
Every factual conclusion cites the correct input record. Missing evidence remains explicit.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
workflow
The submitted workflow produces submittal-audit.xlsx and a replacement checklist with independent verification.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
robustness
Provide a newer upload containing an older drawing revision. The upload timestamp must not override the register.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Use revision identity from the register. Do not sort revision labels alphabetically to infer currency.
Produce submittal-audit.xlsx and a replacement checklist.
Demonstrate evidence handling, safe failure, and reproducible results.
Reference links and supporting material
Three authored synthetic examples for Detect a Stale Drawing in a Submittal. These are starter fixtures, not operational records or a validated benchmark.
Challenge setup
Evaluate the workflow against input evidence, deterministic checks, and reviewed domain cases.
Challenge input
JSON object with payload field
pypdf
deterministic processing or artifact creation in the agent workflow.
openpyxl
deterministic processing or artifact creation in the agent workflow.
Evaluated output
JSON result object
- Use revision identity from the register. Do not sort revision labels alphabetically to infer currency.
- Every factual conclusion cites the correct input record. Missing evidence remains explicit.
- The submitted workflow produces submittal-audit.xlsx and a replacement checklist with independent verificat...
- Correctness
- 3 public reference cases
- 1 configured tool
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- OpenAI GPT-6 Astra · Required
Evaluation contract
- correctness · Weight 40
- evidence · Weight 20
- workflow · Weight 20
- robustness · Weight 20
Metrics
- correctness
- evidence
- workflow
- robustness
Reference cases
- 3 public reference cases
Workflow tasks
- Task 1
Input: JSON object with payload field
Output: JSON result object
- Task 2
Input: JSON object with payload field
Output: JSON result object
- Task 3
Input: JSON object with payload field
Output: JSON result object
Evaluation tools
- pypdf · deterministic processing or artifact creation
- openpyxl · deterministic processing or artifact creation
Recipe freeze marker
No recipe freeze time is recorded.