Docsferatu: An agent that drafts docs PRs while you sleep
Comments on “Docsferatu: An agent that drafts docs PRs while you sleep”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
Docsferatu is a side project that watches an SDK repo for pull requests, uses a classifier to decide whether docs need updating, and if so has Claude draft a docs PR in an isolated sandbox.
Why it ranked
No. 27 of 43 with a score of 0.56. The panel rated it highest on new information (0.90) and technical depth (0.87), lowest on significance (0.36). Placed No. 27, No. 40 by score alone: the fifth Programming & dev tools story of the day (−0.040). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Detailed engineering writeup of an AI docs-drafting agent built on durable execution, with architecture and sandboxing specifics. Useful to builders, but a personal side project with modest broad significance and some self-promotional tone.
- AI judgeJudge 3gpt-5.6-luna
- Detailed first-person engineering write-up covering durable execution, event orchestration, sandbox isolation, prompt-injection defenses, retries, snapshots, and real runs. Highly actionable, though still an early, small-scale project with limited empirical evaluation.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.56total under rubric v2.3
- Significance0.36
- Reader interest0.70
- Audience breadth0.42
- Practical relevance0.68
- Substantiated0.70
- Technical depth0.87
- New information0.90
- Counts against: Promotional0.11
- Counts against: Overclaim0.15
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.41 | 0.20 | 0.48 | 0.36 | 0.28 | +0.24 | +0.087 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.74 | 0.50 | 0.86 | 0.70 | 0.36 | +0.26 | +0.182 |
| Audience breadthHow many people who build technology does it concern? | 0.53 | 0.30 | 0.42 | 0.42 | 0.23 | +0.13 | +0.054 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.77 | 0.50 | 0.78 | 0.68 | 0.28 | +0.22 | +0.150 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.77 | 0.50 | 0.82 | 0.70 | 0.32 | +0.05 | +0.035 |
| Technical depthHow technically substantive is it? | 0.97 | 0.80 | 0.84 | 0.87 | 0.17 | +0.05 | +0.044 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.85 | 0.90 | 0.96 | 0.90 | 0.11 | +0.05 | +0.045 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.06 | 0.20 | 0.08 | 0.11 | 0.14 | −0.19 | −0.022 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.18 | 0.20 | 0.08 | 0.15 | 0.12 | −0.11 | −0.017 |
| Sum of contributions | +0.559 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.559 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-02/27/"><img src="https://rubric.hrih.si/badge/2026-11-02/27.svg" alt="Ranked #27 on Rubric, 2 Nov 2026" width="238" height="28"></a>Claim check
Claim check: 2 matches source · 1 no source given · 5 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
Matches sourceCausal conclusion
With Inngest durable execution, when a function is called again after a crash, redeploy or retry, finished steps return their saved results instead of running again.
The cited source says the same thing.
The article says “finished steps return their saved results instead of running again, and execution picks up after the last one that completed”
The source says “If the run waits or a later step fails, Inngest uses saved results for completed steps without running their callbacks again”
Source cited by the article: inngest.com/docs/durable-execution/durable-workflows
Matches sourceNumber
Inngest's debounce waits until a PR has been quiet for five minutes, then runs once with the latest event.
The cited source says the same thing.
The article says “waits until a PR has been quiet for five minutes, then runs once with the latest event”
The source says “This TypeScript v4 function waits five minutes after the most recent company/updated event for an account. Each new event for that account restarts the five-minute wait”
Source cited by the article: inngest.com/docs/durable-execution/flow-control/debounce
Couldn’t checkCausal conclusion
Inngest's step.sleep saves a wake-up time and ends the invocation, so the Worker is not held open or billed while waiting.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “It saves "wake me in 60 seconds" and the invocation ends. Nothing is held open, and the Worker isn't running (or billing) while the agent thinks”
The source says No passage could be compared.
Source cited by the article: inngest.com/docs/reference/typescript/v4/functions/step-sleep
No source givenNumber
In the author's tests, building a dependency snapshot took about 2.5 minutes, and a sandbox started from it was ready in about 6 seconds.
The article gives no link or citation for it.
The article says “In my tests, building one took about 2.5 minutes, and a sandbox started from it was ready in about 6 seconds”
The source says No source to compare with.
Couldn’t checkDate
For inngest-js#1837, the docs-sync label was added at 19:34:03 UTC, the Worker opened draft website#1995 at 19:40:56 UTC, and the SDK PR merged at 19:55:08 UTC.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “19:40:56 UTC — the Worker makes one commit and opens draft website#1995”
The source says No passage could be compared.
Source cited by the article: github.com/inngest/website/pull/1995
Couldn’t checkNumber
For inngest-js#1834, the draft website#1994 touched four pages under pages/docs/labs/ci/ and opened about 6 minutes after the label was added (18:33:42 to 18:39:49 UTC).
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “touched four pages under pages/docs/labs/ci/ and opened about 6 minutes after the label went on (18:33:42 → 18:39:49 UTC”
The source says No passage could be compared.
Source cited by the article: github.com/inngest/website/pull/1994
Couldn’t checkCausal conclusion
The agent's first real draft, on inngest-js#1830, failed because the default-deny hook denied the Agent SDK's internal StructuredOutput tool, which returns the final JSON answer.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “Turns out the Agent SDK hands back its final JSON answer through an internal tool called StructuredOutput. Because my hook had never heard of it, it denied it”
The source says No passage could be compared.
Source cited by the article: github.com/inngest/inngest-js/pull/1830
Couldn’t checkNumber
For inngest-js#1830, a behavior fix about sandbox cleanup after an error, the agent decided no docs changes were needed and opened no PR.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “With the hook fixed, the agent looked for a while then decided nothing needed to change. No PR, no busywork for a reviewer”
The source says No passage could be compared.
Source cited by the article: github.com/inngest/inngest-js/pull/1830
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 2 pts, not on front page
Recorded for context only, checked 2 Nov 2026, 05:46 UTC. It played no part in the ranking.