Claude Code regularly sends each background agent’s transcript to its model
A third-party gist reports that Claude Code, since version 2.1.232, sends each background agent's transcript to the agent's model about every 30 seconds to generate a 3-5 word status summary, and that this summary requests accounted for roughly 16-20% of measured spend in one user's OpenTelemetry data. The author provides a Bun preload script that wraps setTimeout/setInterval and suppresses the summary timer by matching source-text anchors, reporting zero agent_summary requests afterward on Linux with version 2.1.301. The author notes the approach fails open after updates that change those strings and that Windows and macOS are untested.
Open original source gist.github.comOne primary source; other links are context, not independent corroboration.
- Primary sourceHacker News · AI
- Discovery linkHacker News
- Linked in sourceDRIM
- Linked in source#86879
- Linked in source#104358
A community post points to the original link. Discussion and upvotes do not establish accuracy.
Open full article view →- Community discussion+5
- Fetched source page+30
- No editor link review+0
- Community link+0
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
A third-party gist reports that Claude Code, since version 2.1.232, sends each background agent's transcript to the agent's model about every 30 seconds to generate a 3-5 word status summary, and that this summary requests accounted for roughly 16-20% of measured spend in one user's OpenTelemetry data. The author provides a Bun preload script that wraps setTimeout/setInterval and suppresses the summary timer by matching source-text anchors, reporting zero agent_summary requests afterward on Linux with version 2.1.301. The author notes the approach fails open after updates that change those strings and that Windows and macOS are untested.
A READER'S EVIDENCE CHECKHow to read this signalSix questions about claims, sources, risks, and open checks
This is a signal-level guide until claim-level evidence is available. “Not recorded” means the evidence has not been assessed here; it does not mean a result is false. How we check AI news →
- 01Who is making each claim?
The claimant and whether each claim was measured or asserted are not recorded in this public brief.
- 02What was actually tested?
Not recorded per claim. Check the source for sample size, tasks, model and benchmark versions, and reproducibility.
- 03Where are the sources?
Original link: gist.github.com. Brief basis: fetched source page. Primary page date and quote locations are not established in this panel.
- 04What could mislead?
Source claims not independently checked; the cost figures are Claude Code's own per-request estimates from one user's telemetry, and the workaround is tested only on Linux.
- 05What is at risk if wrong?
Not assessed per claim. Consider safety, privacy, cost, and adoption decisions before relying on the result.
- 06What would change the view?
No claim-specific check is recorded. An inspectable primary method or independent replication could change the assessment.
Key findings from the stored brief
- Claude Code sends each background agent's transcript to its model roughly every 30 seconds to produce a 3-5 word status line, per the author.
- In one OpenTelemetry analysis, agent_summary requests accounted for 16.4% of Claude Code spend versus 15.7% for the main conversation.
- The author's Bun preload suppresses the summary timer by matching source text, and reports zero agent_summary requests on Linux with 2.1.301.
- The author states the documented setting only enables summaries, and the interval is server-controlled with a 30-240 s clamp, so no setting disables it.
Source claims not independently checked; the cost figures are Claude Code's own per-request estimates from one user's telemetry, and the workaround is tested only on Linux.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
Related archive coverage
These relationships record a citation or topic overlap, not confirmation of the finding.
A community post points to the original link. Discussion and upvotes do not establish accuracy.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
A publication revision history is not stored for this signal. New engagement snapshots do not mean the source itself changed.
Why this may matter
If accurate, background summary calls can consume a large share of tokens and rate-limit headroom without users' knowledge. The post also illustrates how a vendor-side default can be changed by update and how users may need to verify behavior after each release.
Limits and open questions
Source claims not independently checked; the cost figures are Claude Code's own per-request estimates from one user's telemetry, and the workaround is tested only on Linux.
Organization-specific financial, security, and operational effects require the organization's own context and review.
Discussion
Comments are reviewed before publication. Discuss the evidence, not the person.
No published comments yet.
Reader discussion opens with accounts. You can read the original source and published comments now.