Write the problem hypothesis
For [user], the agent loses or retrieves [information] incorrectly across [scope], causing [wrong action or repeated work].
You can name one user, one situation and one measurable consequence without proposing a feature.Long-Context & Complex Reasoning Coding Evaluation Dataset
Evidence observed in google-gemini/gemini-cli, a Coding project.
80 comments14 positive reactions171 days openProject Radar 97
priority/p2area/platform🔒 maintainer onlystatus/bot-triagedkind/enhancement
Start free validation sprint4 guided steps · private notes · cloud sync
SOURCE EVIDENCE
Read original GitHub Issue ↗Start with what users actually said
Read the complete discussion before interpreting the problem. Note repeated use cases, workarounds, constraints and the people asking for the change.
Interview five affected users
- Which information must persist?
- For how long and within what scope?
- Show the last wrong or missing recall.
- What data must never cross boundaries?
- How do you correct memory today?
Run the smallest experiment
Test one constrained memory scope with a small real dataset and a written expected-recall benchmark.
The system retrieves the right evidence across repeated sessions without leaking or inventing context.Make a build decision
- Build: repeated pain and active commitment
- Narrow: pain is real but the audience or job differs
- Stop: weak frequency or no behavioral proof
Why this brief exists
Information has value only when it changes action. This page turns one public signal into a bounded validation exercise. It is a research aid, not proof of demand, investment advice or a product recommendation.