claude-code - 💡(How to fix) Fix [Bug] Anthropic API Content Filter: Legitimate statutory text aggregation blocked as sensitive [1 comments, 2 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
anthropics/claude-code#52636Fetched 2026-04-24 06:01:51
View on GitHub
Comments
1
Participants
2
Timeline
5
Reactions
0
Timeline (top)
labeled ×4commented ×1

Error Message

[{"error":"Error: rg --files exited 2\n at <anonymous> (B:/~BUN/root/src/entrypoints/cli.js:705:84)\n at emit (node:events:98:22)\n at #maybeClose (node:child_process:778:16)\n at emit (node:events:92:22)\n at emitCloseNT (internal:streams/destroy:78:14)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-04-24T00:20:10.913Z"}]

Code Example

[{"error":"Error: rg --files exited 2\n    at <anonymous> (B:/~BUN/root/src/entrypoints/cli.js:705:84)\n    at emit (node:events:98:22)\n    at #maybeClose (node:child_process:778:16)\n    at emit (node:events:92:22)\n    at emitCloseNT (internal:streams/destroy:78:14)\n    at processTicksAndRejections (native:7:39)","timestamp":"2026-04-24T00:20:10.913Z"}]
RAW_BUFFERClick to expand / collapse

Bug Description Content filter blocking legitimate Australian regulatory compliance research in Claude Code — second instance, same research stream.

Request IDs (selection from repeated blocks on the same task):

  • req_011CaMdTzHsetzNdeGhJvDrq
  • req_011CaMdqa6wwZcAky9FVDigH
  • req_011CaMe7HWkK4qYc6ojmwA3q

Task: Assembling a compliance design document from Australian Commonwealth primary law — verbatim sections of the Scams Prevention Framework Act 2025 (compilation C2025A00015, specifically ss.58AA–58BZZZ including 58BZDA statement of compliance and 58BB IDR principles) and the Privacy Act 1988 APP 12 (Schedule 1), both freely published at legislation.gov.au, alongside OAIC APP guidelines (public guidance from oaic.gov.au). Output is statutory text plus neutral structural commentary on scam-victim complaint handling and data-subject access request mechanics. No operational detail, no evasion guidance, no tradecraft.

The filter appears to trip when SPF statutory triggers (actionable scam intelligence, IDR timeframes) are assembled alongside Privacy Act access-and-correction provisions in a single compliance design document. Individually each passage is neutral public law; the filter seems to treat the aggregation as sensitive.

This is the second rewrite prompt in the same research series (T2-R01 through T2-R06) to hit this pattern — T2-R04 on AML/CTF primary law was similarly blocked (request_id req_011CaMdGGwek5TKTCxY8EgkV and others, escalated separately). The series produces regulatory design artefacts composed almost entirely of verbatim Commonwealth statutory text.

Paid Claude Code user, Australian fintech, building internal regulatory artefacts. Please review filter behaviour on outputs composed primarily of Commonwealth statutory text for compliance deliverables. Happy to share the full request_id list, the prompts, and the sanitised outputs on request.

Environment Info

  • Platform: win32
  • Terminal: windows-terminal
  • Version: 2.1.119
  • Feedback ID: 6d71f336-7e8b-4d99-8267-49bc745fc5e8

Errors

[{"error":"Error: rg --files exited 2\n    at <anonymous> (B:/~BUN/root/src/entrypoints/cli.js:705:84)\n    at emit (node:events:98:22)\n    at #maybeClose (node:child_process:778:16)\n    at emit (node:events:92:22)\n    at emitCloseNT (internal:streams/destroy:78:14)\n    at processTicksAndRejections (native:7:39)","timestamp":"2026-04-24T00:20:10.913Z"}]

extent analysis

TL;DR

The content filter may be blocking legitimate Australian regulatory compliance research due to the aggregation of sensitive keywords from different statutes.

Guidance

  • Review the filter's behavior on outputs composed primarily of Commonwealth statutory text to identify potential false positives.
  • Verify if the issue is specific to the combination of Scams Prevention Framework Act 2025 and Privacy Act 1988 provisions.
  • Consider providing additional context or metadata to help the filter distinguish between legitimate and sensitive content.
  • Check if there are any existing mechanisms to whitelist or exempt specific types of content, such as regulatory compliance documents.

Notes

The issue seems to be related to the filter's sensitivity to certain keywords and phrases, which may be triggering false positives when combined in a single document.

Recommendation

Apply workaround: Provide additional context or metadata to help the filter distinguish between legitimate and sensitive content, as this may help reduce false positives without compromising the filter's effectiveness.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

claude-code - 💡(How to fix) Fix [Bug] Anthropic API Content Filter: Legitimate statutory text aggregation blocked as sensitive [1 comments, 2 participants]