Trilemma: Recursive tasks, concurrency cap, no deadlock in XL agentic systems
Comments on “Trilemma: Recursive tasks, concurrency cap, no deadlock in XL agentic systems”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
CocoIndex describes a rule for task trees under a concurrency cap: a waiting parent lends its slot to one child at a time, which avoids deadlock and keeps in-flight tasks at most cap times tree depth. It says releasing the slot at the first child, its approach through 1.0.25, let an indexer reach about 15,000 file components in flight and run out of memory. In its own 2,000-file benchmark with a cap of 16, in-flight components fell from over 1,100 to 16.
Why it ranked
No. 21 of 142 with a score of 0.68. The panel rated it highest on technical depth (0.97) and substantiated (0.93), lowest on significance (0.58). Placed No. 21, No. 25 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, reproducible engineering writeup on concurrency limits in recursive task trees with runnable code and benchmarks; technical and practical, but niche and partly promotional for the author's engine.
- AI judgeJudge 3gpt-5.6-luna
- A deeply substantiated engineering write-up with reproducible models, benchmarks, implementation details, proofs, and practical guidance for bounding recursive async workloads without deadlock or memory blowups. It promotes its engine somewhat but provides substantial transferable knowledge.
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
0.68total under rubric v2.3
- Significance0.58
- Reader interest0.79
- Audience breadth0.64
- Practical relevance0.82
- Substantiated0.93
- Technical depth0.97
- New information0.87
- Counts against: Promotional0.23
- Counts against: Overclaim0.20
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? | 0.66 | 0.35 | 0.72 | 0.58 | 0.37 | +0.24 | +0.138 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.77 | 0.70 | 0.90 | 0.79 | 0.20 | +0.26 | +0.205 |
| Audience breadthHow many people who build technology does it concern? | 0.79 | 0.40 | 0.72 | 0.64 | 0.39 | +0.13 | +0.083 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.95 | 0.60 | 0.90 | 0.82 | 0.35 | +0.22 | +0.180 |
| 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. | 0.99 | 0.85 | 0.95 | 0.93 | 0.14 | +0.05 | +0.047 |
| Technical depthHow technically substantive is it? | 1.00 | 0.92 | 0.98 | 0.97 | 0.08 | +0.05 | +0.048 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.87 | 0.85 | 0.90 | 0.87 | 0.05 | +0.05 | +0.044 |
| 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. | 0.08 | 0.30 | 0.30 | 0.23 | 0.22 | −0.19 | −0.043 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.21 | 0.20 | 0.20 | 0.20 | 0.01 | −0.11 | −0.022 |
| Sum of contributions | +0.679 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.679 | ||||||
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/2026-11-18/21/"><img src="https://rubric.hrih.si/badge/2026-11-18/21.svg" alt="Ranked #21 on Rubric, 18 Nov 2026" width="245" height="28"></a>Claim check
Claim check: 3 matches source · 4 no source given · 1 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
Matches sourceNumber
Claude Code caps its sub-agent tree at 3 levels and 20 running sub-agents by default, and refuses a new one at the cap.
The cited source says the same thing.
The article says “Claude Code caps its sub-agent tree at 3 levels and 20 running sub-agents by default, and refuses a new one at the cap”
The source says “By default, when 20 subagents are running in a session, spawning another with the Agent tool fails with Concurrent subagent limit reached, and the error tells Claude not to retry”
Source cited by the article: code.claude.com/docs/en/sub-agents
Matches sourceNumber
An indexer built on CocoIndex was killed for running out of memory with about 15,000 file components in flight, despite a 256-file application gate and an engine component cap.
The cited source says the same thing.
The article says “Yet when it died, about 15,000 file components were in flight, and each one held a file’s contents”
The source says “one indexer was OOM-killed with ~15k file components in flight while an application gate admitted 256 of them”
Source cited by the article: github.com/cocoindex-io/cocoindex/pull/2486
Couldn’t checkVersion
CocoIndex released a component's slot on its first child mount from the day the component limit was added through 1.0.25.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “This was CocoIndex’s rule from the day the component limit was added (#1678) through 1.0.25: a component gave its slot back on its first child mount”
The source says No passage could be compared.
Source cited by the article: github.com/cocoindex-io/cocoindex/pull/1678
No source givenBenchmark result
On the engine benchmark with a cap of 16, CocoIndex 1.0.25 reached over 1,100 file components in flight while 1.1.1 with slot lending stayed at 16.
The article gives no link or citation for it.
The article says “reached over 1,100 file components in flight under a cap of 16 with this rule (CocoIndex 1.0.25). With lending (1.1.1), it stays at 16”
The source says No source to compare with.
No source givenNumber
Peak memory in the engine benchmark was 167-168 MB on 1.0.25 versus 33 MB on 1.1.1.
The article gives no link or citation for it.
The article says “1.1.1 (slot lending), run 1 16 33 MB”
The source says No source to compare with.
No source givenVersion
CocoIndex has used slot lending since version 1.0.26.
The article gives no link or citation for it.
The article says “CocoIndex has done this since 1.0.26”
The source says No source to compare with.
No source givenFirst, only or fastest
Lending is the only design that gives a hard cap on executing tasks, a bound on tasks in flight and no need to know the levels.
The article gives no link or citation for it.
The article says “Lending is the only row with a hard cap on executing tasks, a bound on tasks in flight and no need to know the levels”
The source says No source to compare with.
Matches sourceQuote
Python's ThreadPoolExecutor docs warn that deadlocks can occur when a callable associated with a Future waits on the results of another Future.
The cited source says the same thing.
The article says “Deadlocks can occur when the callable associated with a Future waits on the results of another Future”
The source says “Deadlocks can occur when the callable associated with a Future waits on the results of another Future”
Source cited by the article: docs.python.org/3/library/concurrent.futures.html
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 5 pts, not on front page
Recorded for context only, checked 18 Nov 2026, 05:47 UTC. It played no part in the ranking.