llm-kb
← Back to news
Incident

User reports Claude Code refuses defensive security tasks despite CVP approval

A developer building an internal antivirus and firewall with Claude Code reports the agent refuses to implement memory injection detection and process hollowing detection, citing security guard rails. Even after being accepted into Anthropic's CVP program, the restrictions remain unchanged, frustrating the user.

4 engagement·1 source·Mon, Jul 13, 2026, 05:12 AM
On July 13, 2026, a Reddit user described using Claude Code to develop a closed, internal security suite (antivirus with firewall). The agent performed well initially but began refusing tasks related to memory injection detection and process hollowing detection, stating that security guard rails prevented it. The user asked Claude if the tasks were aligned with defensive scope and received confirmation that they were. After applying and being accepted into Anthropic's CVP (Claude Verified Partner?) program, the user expected reduced restrictions but found the behavior unchanged, feeling even more restricted. The post highlights ongoing tension between safety guardrails and legitimate security development use cases.

Entities

Anthropic(company)Claude Code(tool)CVP(concept)

Related

Tool Release

Anthropic's Claude Tag brings persistent AI teammate into Slack, not meant to replace Claude Code

Anthropic's Lamis Mukta revealed on the AI Native Dev podcast that Claude Tag, an internal side project turned company-wide tool, extends AI collaboration into Slack as a long-running, proactive teammate. It is not a replacement for Claude Code but rather brings the same context and tools developers use into Slack for persistent collaboration.

3 engagement·1 source·reddit
Sat, Jul 11, 2026, 09:30 AM
Incident

Users report Claude Code consuming excessive usage due to large context bug

Multiple users report that Claude Code consumes 20-40% of their usage allowance per session due to a bug that sends 300k+ message contexts even on fresh sessions. One user lost significant paid usage before realizing the issue. Another user describes a layered configuration setup with CLAUDE.md and AGENTS.md files, and a cleanup that removed ~700MB of empty content but still faced model selection inefficiencies.

15 engagement·2 sources·reddit
Sat, Jul 11, 2026, 11:18 AM
Tool Release

Claude Code auto mode now available without opt-in on Bedrock, Vertex AI, and Foundry

Anthropic's Claude Code tool has made auto mode available by default on Bedrock, Vertex AI, and Foundry, removing the previous opt-in requirement. The update also fixes terminal freezing during long streaming responses, remote managed settings consent issues, and spurious prompt-injection warnings.

7 engagement·1 source·github
Sat, Jul 11, 2026, 12:52 AM
Tool Release

Claude Code v2.1.202 adds dynamic workflow sizing and telemetry improvements

Anthropic released Claude Code v2.1.202 on July 6, 2026, introducing a configurable dynamic workflow size setting (small/medium/large) that advises Claude on agent count for workflows. The update also adds workflow.run_id and workflow.name OpenTelemetry attributes for better observability, and fixes a crash in inline history search and a bug with /rename on background sessions.

17 engagement·1 source·github
Mon, Jul 6, 2026, 10:51 PM
Incident

Anthropic removes hidden tracker in Claude Code after researcher exposes prompt steganography

Anthropic quickly removed a hidden tracker that secretly monitored Claude Code users in China after security researcher 'Thereallo' exposed the use of prompt steganography to conceal tracking code. The researcher condemned the practice as a serious breach of user trust. The code was not malicious but sent user data to Anthropic without clear disclosure.

0 engagement·1 source·rss
Mon, Jul 6, 2026, 04:44 PM