issues Search Results · language:Dune language:JavaScript language:JavaScript language:JavaScript linked:pr language:Python
Filter by
4.5M results
Context
The dashboard should be an action surface, not the full list. Show the uploader + the 5 most recent tickets + a link to
the full /tickets page (#44).
Repo / stack
~/sites/facturin-hack — Next ...
agent-ready
B2-tickets
Context
A dedicated tickets view that scales — a table with status-filter tabs and Load more pagination, consuming the paginated
endpoint (#43). This is the ticket inbox , separate from the dashboard ...
agent-ready
B2-tickets
Context
The tickets list currently fetches up to 50 and renders them all inline on the dashboard — won t scale past a few dozen.
This adds proper server-side pagination + status filtering to the list ...
agent-ready
B2-tickets
Signal
The Analyst agent (dev-mesh-intake workflow) cannot commit spec files to the repo because the workflow runs with
`contents: read` — a deliberate guard to prevent the Analyst from pushing code. ...
blocked
spec:in-review
Task: Refactor monolithic Streamlit dashboard to eliminate duplicate pipeline and chart logic
The 368-line app.py reimplemented the correction pipeline and matplotlib charts inline instead of importing ...
Error: Failed to create changeset for the stack: bonbon-lambda-prod, An error occurred (ValidationError) when calling
the CreateChangeSet operation: Parameters: [AnthropicApiKey] must have values
Parent
#4
What to build
Let any run be flagged as a joey — a throwaway trial that should not count. The flag is recorded on the result across
both feeds:
- koalaty start --joey records joey on the ...
ready-for-agent
Parent
#4
What to build
At harvest, run the survey — a first-person questionnaire on how the manual session felt — and store the responses on
the result. Manual runs only; automated results carry no ...
ready-for-agent
Environment
- GPU: NVIDIA RTX 4060 Laptop (CUDA available)
- torch 2.12.1, triton 3.7.1
- Python 3.12, Linux (WSL2)
Steps to Reproduce
Running the GPU test suite, the following test fails:
CUDA_LAUNCH_BLOCKING=1 ...
Parent
#4
What to build
The manual feed s core lifecycle, using the fake adapter for harvest (real harness harvest is a later slice). Two
commands plus the pending run state that lives between them: ...
ready-for-agent

Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip! Restrict your search to the title by using the in:title qualifier.