Replay three AI agent approval failures in five seconds, in Python
Quiet launch
Beat 32% of Show HN launches that day
#115 of 182 on Nov 5
#19 of 35 AI Agents launches on Show HN on Nov 5. 4.7% of AI Agents launches (8 of 172) reached 10+ points in the last 7 days. See AI Agents →
points · traction gained Nov 11, 10:42 UTC → Nov 12, 10:31 UTC · % shown only when the previous traction was 5+
Traction frozen — tracking ended at 160h after listing. Show HN votes are re-read for ~168h.
comments
listed Nov 5, 18:17 UTC
Traction (points) over time
Key statistics
- Traction
- 2points
- Comments
- 0
- Rank in AGNT this week
- —not listed in the last 7 days
- Rank overall this week
- —
- Sector hit rate
- 4.7%8/172 · 7D
- Traction line
- Below10+ points
- First reading
- 2Nov 6, 17:32 UTC
- Last reading
- 2Nov 12, 10:31 UTC
- Readings
- 7over 160h
Get the launches that broke through, every Tuesday.
8 of 172 AI Agents launches on Show HN reached 10+ points over the last 7 days. The weekly brief shows what they did differently, so you can compare Replay three AI agent approval failures in five seconds, in Python.
Free, every Tuesday. Unsubscribe anytime.
Top on Nov 5 · 182 Show HN launches that day, by points
- #1FirenookFirebase emulators in Rust: 40× faster writes, 10× less RAM176 points
- #2Rivet (YC W23)Scale and Orchestrate Agents on S3122 points
- #3Stroqa local audit of what your coding agent did and why96 points
- #4WaymarkA workflow engine that compiles jobs instead of replaying88 points
- #5Spindlea full-featured, type-safe web framework for OCaml 564 points
More in AI Agents · top Show HN launches, last 7 days
- Get your coding agent off ~2019 JavaScript, onto modern web primitives439 points▲ +10
- AgentRunopen-source, Self-improving Managed Agents103 points▲ +103
- Klad, a paint app your coding agent can drive (source-available)83 points▲ +4
- Pomerado (YC F26)Turn any website into an MCP for agents61 points▲ +6
- RecalcSpreadsheet where cells can be Claude Code or Codex terminals37 points0
Grades · each graded against other AI Agents launches from the same source
Momentum
——
Traction gained in the latest daily interval
Not re-read in the latest run, so no recent gain
Velocity
C0.04 points/h
Votes per hour since listing (age capped at 48h)
Better than 57% · vs 746 AI Agents · Show HN launches, last 30 days
Engagement
——
Comments per 10 points
Needs 10+ points; below that the ratio is mostly noise
Staying Power
——
Votes at last reading ÷ votes at first reading
Needs 10+ points and readings from two runs
Launch-day race · #115 of 182 Show HN launches on 2026-11-05
- #113Keynate2 points
- #114Lunar2 points
- #115Replay three AI agent approval failures in five seconds, in Python2 points
- #116Resilient AI Worker Queue with Redis, BullMQ, and Budget Kill Switches2 points
- #117Magertron2 points
Maker? Add your website and we will link it from this page.
Discussion
No comments yet. Open on Hacker News →
Keep going
- #116 of 182 on Nov 5Resilient AI Worker Queue with Redis, BullMQ, and Budget Kill Switches →
- Case studyHow HackerBooks reached 1,861 points on Show HN with a five-word pitch →
- Case studyHow "Steal My Bitcoin" turned a two-word dare into the #1 Show HN of the week →
- CategoryAI Agents: supply, hit rate, top launches →
Every Tuesday: the launches that broke through and the categories heating up. Get the free weekly brief →
Data as of Nov 13, 2026, 9:22 AM UTC. Sector comes from keyword rules on the title and tagline and can be wrong. How it works