Web search: curate allowlist/denylist
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 5
- Forks
- 10
- Avg merge
- 2d 20h
- Merged PRs (30d)
- 17
Description
Follow-up on #664
With a very limited allowlist, the assistant may try to search for external information several times, trying to find relevant urls it is allowed to open. For example, with just OpenFn documentation in the allowlist, and the user question "What is the latest format for tracked entity attributes according to the DHIS2 API? I need the exact field names", the assistant is likely to spend all of its allowed tool uses trying to find the information within OpenFn urls.
We should optimise the allowlist and prompts/tool use descriptions to avoid having the model stall at search attempts.
Create a couple of acceptance tests, but we should also test the internal tool use behaviour and not just the answer.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by tracing the web-search allowlist and denylist behavior, then review the prompts and tool-use descriptions that govern repeated search attempts. Add the requested acceptance tests and exercise internal tool-use behavior as well as the final answer, confirming that restricted searches do not consume all allowed tool uses.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- jupyter-notebook
- Domain
- ai, search, testing
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100