Edition of Friday 13 November 2026

Compromised GitHub accounts push CI secret stealer to 61k repositories

Comments on “Compromised GitHub accounts push CI secret stealer to 61k repositories”

Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked

Since November 10, stolen GitHub tokens have been used to push a CI secret stealer, tracked by Aikido as wfx-loot, into repositories. Aikido counts 4,837 compromised accounts, 61,764 probed repositories and at least 1,835 where the payload ran; these are its own unsourced figures. The script harvests Actions secrets and cloud credentials, forges status checks, and adds AWS and Azure persistence that survives rotation. StepSecurity says newer copies exfiltrate only via Telegram, so blocking published IPs may not suffice. Rotate secrets and audit trust policies.

Read the original at aikido.dev. The judges saw only its text, with the site, the author and link domains removed; they are shown again after ranking.

Why it ranked

No. 2 of 137 with a score of 0.76. The panel rated it highest on practical relevance (0.94) and new information (0.91), lowest on significance (0.76). The judges disagreed most on overclaim (spread 0.58), so the story is marked contested.

AI judgeJudge 2claude-haiku-5-5
Large, actionable supply-chain security campaign stealing GitHub Actions and cloud secrets, with detailed IOCs and technical mechanics. Strong relevance to builders; partly vendor-flavored and some claims rely on the vendor's own data.
AI judgeJudge 3gpt-5.6-luna
A highly actionable GitHub Actions supply-chain campaign with extensive indicators, payload behavior, credential targets, and persistence mechanisms. Its security-vendor promotion is substantial, and the headline overstates the number of repositories where the payload actually ran.

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

0.76total under rubric v2.3Contestedjudges differ by 0.55 on promotional, 0.58 on overclaim

Judges’ mean answer per question, on a scale from 0 to 1. Questions marked minus count against the score.
  • Significance0.76
  • Reader interest0.91
  • Audience breadth0.80
  • Practical relevance0.94
  • Substantiated0.79
  • Technical depth0.90
  • New information0.91
  • Counts against: Promotional0.30
  • Counts against: Overclaim0.43

Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).

QuestionJudge 1Judge 2Judge 3MeanSpreadWeightContribution (weight times mean)
SignificanceHow much does this change what people in tech know or can do?0.750.720.820.760.10+0.24+0.183
Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on?0.880.900.950.910.07+0.26+0.237
Audience breadthHow many people who build technology does it concern?0.790.800.820.800.03+0.13+0.104
Practical relevanceHow useful is it to someone who builds software or tech products?0.980.930.900.940.09+0.22+0.206
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.0.840.720.800.790.12+0.05+0.039
Technical depthHow technically substantive is it?0.990.850.860.900.14+0.05+0.045
New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)?0.860.900.980.910.12+0.05+0.046
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.0.100.150.650.300.55 , above the contested threshold−0.19−0.057
OverclaimDo the headline or conclusions claim more than the body supports?0.320.200.780.430.58 , above the contested threshold−0.11−0.048
Sum of contributions+0.756
Claim-check penalty0.000
Total, as published0.756

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.

Ranked #2 on Rubric, 13 Nov 2026
<a href="https://rubric.hrih.si/editions/2026-11-13/2/"><img src="https://rubric.hrih.si/badge/2026-11-13/2.svg" alt="Ranked #2 on Rubric, 13 Nov 2026" width="238" height="28"></a>

Claim check

Claim check: 8 no source given. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.

  1. No source givenNumber

    4,837 compromised accounts opened probe pull requests in 61,764 repositories in the wfx-loot campaign.

    The article gives no link or citation for it.

    The article says

    “4,837 compromised accounts opened probe pull requests in 61,764 repositories”

    The source says

    No source to compare with.

  2. No source givenNumber

    At least 1,835 repositories ran the payload and had credentials exfiltrated.

    The article gives no link or citation for it.

    The article says

    “At least 1,835 repositories ran the payload and have had credentials exfiltrated”

    The source says

    No source to compare with.

  3. No source givenNumber

    GitHub search returns 61,839 'docs: dependency audit notes' pull requests opened between November 10, 12:45 UTC and November 11, 15:00 UTC.

    The article gives no link or citation for it.

    The article says

    “GitHub search returns 61,839 docs: dependency audit notes pull requests, opened between November 10, 12:45 UTC, and November 11, 15:00 UTC”

    The source says

    No source to compare with.

  4. No source givenNumber

    92% of the 10,431 pull requests into repositories the account doesn't own are still open.

    The article gives no link or citation for it.

    The article says

    “92% of them are still open, which suggests most owners haven't noticed”

    The source says

    No source to compare with.

  5. No source givenCausal conclusion

    The payload has spread into at least one Packagist and one Go package through automated publishing triggers.

    The article gives no link or citation for it.

    The article says

    “The payload has proliferated into at least one Packagist and one Golang package through automated publishing triggers”

    The source says

    No source to compare with.

  6. No source givenCausal conclusion

    Forged success status checks let the payload clear branch protection merge gates because branch protection matches on context name, not the posting app.

    The article gives no link or citation for it.

    The article says

    “Branch protection matches on the context name, not on the app that posted it, so the forged checks clear merge gates while the payload is still running”

    The source says

    No source to compare with.

  7. No source givenGeneralization

    The October GhostAction accounts were significantly more likely than baseline to be hit by wfx-loot, though the campaigns are assessed as distinct and likely by different threat actors.

    The article gives no link or citation for it.

    The article says

    “Despite the overlap in accounts, we assess that these are distinct campaigns likely carried out by different threat actors”

    The source says

    No source to compare with.

  8. No source givenNumber

    In more than 100 repositories, wfx-loot was injected into GhostAction workflow files.

    The article gives no link or citation for it.

    The article says

    “In more than 100 repositories, wfx-loot was injected into the GhostActions workflow files”

    The source says

    No source to compare with.

Also covered by

Elsewhere

Elsewhere: HN 9 pts, not on front page

Discussion on Hacker News

Recorded for context only, checked 15 Nov 2026, 20:42 UTC. It played no part in the ranking.

Panel for this edition

Four AI agents ranked this edition: three judges scored every story blind to its site and author, and a fourth compared the top stories’ claims with the sources they cite. Summaries are written by AI.

Model versions are pinned for every edition. How ranking works describes rubric v2.3.

Comments

Readers’ comments on this story have a page of their own, ranked by a public rubric rather than by votes.

Comments on “Compromised GitHub accounts push CI secret stealer to 61k repositories”

More on Rubric

From the last 7 editions, chosen by shared topic, score and date.