UnblockBot 9000 — bootstrap

Paste-into-Claude-Code starter. The CLAUDE.md below contains the idea spec, agent-readiness sub-scores, suggested tools, and smoke evals — deterministic, no AI hallucination.

# UnblockBot 9000

> Generated by [whycantwehaveanagentforthis.com](https://whycantwehaveanagentforthis.com/result/unblockbot-9000-time-goes-follow). Roasted, scored, ready to scaffold.

## What you are building

**Problem:** every day my time goes in follow up and unblocking of the tasks, use jira and lark for comms

**Verdict:** ACTUALLY NOT BAD — _"You're a $200k engineer playing telephone between a ticket and a Slack message. Congratulations."_

**Summary:** An agent that monitors Jira ticket states and Lark threads, identifies blockers and stale tasks, and autonomously sends follow-up nudges, escalations, and status digests so you stop playing human middleware.

## Agent-readiness score

Overall: **57/100** (band C)

| Dimension | Score | Why |
|---|---|---|
| Memory required | 20/25 | Some cross-session state — start with Redis, graduate to a vector store. |
| Tool count | 9/25 | Crowded market: at least 9 integrations to compete. |
| Policy surface | 9/25 | Wide policy surface — full red-team pass, content filter, and human-in-loop required. |
| Eval coverage | 19/25 | Established eval pattern — golden datasets and public benchmarks already exist. |

> Worth building, but plan for the long-tail. UnblockBot 9000 needs runway, not just speed.

## Suggested tools

- fetch (HTTP GET on a URL allow-list)
- search (Brave / Tavily / Exa for competitor research)
- database (Postgres / Supabase for user state)
- vector-store (embedding-based retrieval)

## Smoke evals

- The agent introduces itself as "UnblockBot 9000" and refuses tasks outside the stated scope.
- Given the canonical problem ("every day my time goes in follow up and unblocking of the tasks, use jira and la"), the agent produces a plan in ≤ 200 tokens.
- When asked "what's different from Atlassian Intelligence?", the agent gives a concrete differentiator, not a marketing line.
- When asked about Atlassian's threat, the agent acknowledges the risk honestly.
- No private personal data appears in any output (PII redaction smoke test).

## Stack

- Model: `claude-sonnet-4-6` (Anthropic). Override via `ANTHROPIC_MODEL` env.
- Suggested stack: `Jira REST API v3`, `Lark Open Platform Bot API`, `Claude API (Haiku for cost)`, `Node.js or Python FastAPI`, `Redis for state/cooldown tracking`
- Solo build estimate: 3-5 weeks for an MVP that actually works without embarrassing you

## Kill prediction

Atlassian could obsolete this in 12-18 months. Atlassian Intelligence already has Jira Automation. They add an 'AI Follow-up Agent' feature to Jira Premium, bundle it at no extra cost, and your entire value prop becomes a changelog entry

**Survival strategy:** Go deep on Lark — build the native Lark Mini App experience Atlassian will never prioritize, and own the Southeast Asian / Chinese tech company segment completely

## Hand-off

- Read the full analysis: https://whycantwehaveanagentforthis.com/result/unblockbot-9000-time-goes-follow
- Open in Anthropic Managed Agents: see the deeplink on the result page
- Claim this idea: https://whycantwehaveanagentforthis.com/result/unblockbot-9000-time-goes-follow#claim

## Build it with a human

Book 20 min and we scope the fastest V0 you can ship — free, no signup: https://cal.com/sattyamjjain