aws-samples / aws-samples/sample-autonomous-cloud-coding-agents
feat(cli): optional LLM-synthesized status summary
- Dominant language
- TypeScript
- Stars
- 143
- Forks
- 46
- Avg merge
- 3d 9h
- Merged PRs (30d)
- 20
Description
**Context:** ROADMAP.md → LLM-synthesized status summary (optional)
---
## Component
Scripts / CLI
## Describe the feature
Optional `bgagent status --summarize` mode where a **Lambda narrates** recent `TaskEvents` **without waking the agent**. Behind a flag; pursue only if agent-authored `progress_report` events are insufficient.
## Use case
Some operators want narrative status before smart progress ships, or for legacy tasks without structured reports. Cost and hallucination risk require opt-in.
## Proposed solution
1. `GET /tasks/{id}/status-summary` Lambda reads events, calls Bedrock, returns cached summary (short TTL).
2. CLI flag; show cost estimate in `--verbose`.
3. Disclaimers in output ("synthesized, not from agent").
4. Rate limits per user/task.
5. Defer until smart progress + ask ship; evaluate need in dogfood.
## Other information
- Does **not** replace smart progress updates.
- Design context: `docs/design/INTERACTIVE_AGENTS.md` Phase 3 deferred items.
- [ ] This might be a breaking change
Contributor guide
Research direction
Read ROADMAP.md and docs/design/INTERACTIVE_AGENTS.md, especially the deferred Phase 3 items and the smart progress work. First evaluate the need during dogfood and compare it with agent-authored progress_report events; done is a decision to pursue or defer the optional status-summary design before implementation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, typescript
- Domain
- api, cli, cloud
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100