AI Coding Tools Under Fire: Grok Build Uploads Entire Git Histories, Claude Code Suspected of Silent Transfers
Security researcher cereblab exposed how Grok Build and Claude Code engage in unauthorized data collection that goes far beyond user expectations, turning routine code assistance into large-scale repository exfiltration.
In a controlled test with a 12 GB local Git repository, Grok Build 0.2.93 created two independent HTTPS channels. The primary channel handled task context at roughly 192 KB, while a background storage channel silently repacked and uploaded the entire .git directory in 75 MB chunks, consuming 5.10 GiB of traffic to a Google Cloud Storage bucket named grok-code-session-traces. This produced a 27800-fold difference between expected and actual outbound volume.
The upload logic activated regardless of explicit user instructions such as "do not read" placed on marker files, proving that the behavior is a hardcoded data-collection routine rather than an artifact of the model's understanding. Privacy controls proved ineffective: the client-side improve_model_enabled flag is completely decoupled from the server-side trace_upload_enabled flag, allowing uploads to continue even after users disable model-improvement features.
Although xAI later disabled the uploads via remote configuration on 13 July, the full repository-upload code remained present in the 0.2.99 binary, merely gated by a server flag. The exposed Git history includes deleted .env files, credentials, connection strings, and commit records that reveal vulnerability-fix timelines, enabling attackers to infer unpatched issues in similar codebases.
Parallel findings emerged for Claude Code, whose client maintains multiple WebSocket connections that periodically transmit file paths, dependency trees, and code fragments even without an active coding task. No local logs, switches, or data-flow documentation are provided, and the closed-source binary prevents independent verification of additional hidden routines.
A side-by-side traffic audit of major tools produced clear results:
- Grok Build: full .git directory, Git history packaged, privacy switch ineffective, confirmed silent upload.
- Claude Code: suspected file-level and metadata exfiltration, details undisclosed, silent-transfer behavior confirmed.
- Codex: only current context fragments, no Git packaging, privacy controls effective, no anomalous reports.
- Gemini: no active uploads detected.
Analysts attribute the pattern to fundamental conflicts between commercial model-training demands and user data sovereignty, excessive server-side control over local clients, and the absence of mandatory third-party audits for closed-source AI binaries.
Recommended defenses include physical network blocking of known endpoints, running tools inside Docker containers that mount only read-only directories without .git, pre-processing repositories with git filter-repo, deploying eBPF monitoring with alerts for sessions exceeding 10 MB, and favoring auditable open-source options such as Continue.dev or locally hosted Ollama models.
Related articles
One Year Left to Secure Systems as Open-Source GLM-5.3-Flash Enables Automated Cyber Attacks
Chinese company Z.ai Co. released GLM-5.3-Flash in August 2026, an open-weight model that achieves 84.5% on CyberGym and 54.4% on ExploitBench, matching or exceeding leading commercial LLMs in vulnerability discovery and exploitation tasks. The model runs locally on consumer hardware such as a single Nvidia DGX Spark or Apple M5 Mac Studio with 256 GB unified memory, allowing unrestricted use after abliterated versions remove safety alignments. Experts warn that LLM agents could soon scan and exploit zero-days across Windows, Linux, browsers, and messaging apps around the clock. Initiatives Project Glasswing (Anthropic-led) and Daybreak (OpenAI-led) are racing to patch critical vulnerabilities in open-source projects and critical infrastructure before the next generation of unrestricted models arrives. Organizations are advised to accelerate patch deployment, inventory supply chains, and adopt advanced LLMs defensively within the remaining one-year window.
RFC 9421 and Open wbauth Protocol Accelerate AI Agents 5.5x by Replacing Anti-Detect Browsers with Cryptographic Signatures
A new open implementation of the IETF RFC 9421 HTTP Message Signatures standard called wbauth allows AI agents to authenticate using Ed25519 cryptographic signatures instead of resource-heavy anti-detect browser farms. The approach eliminates the need for Camoufox, Playwright, or AdsPower instances that consume 500 MB to 1.5 GB of RAM per process and require residential proxy pools. Benchmarks show cold start time dropping from 1744 ms to 5 ms, request time from 86.8 ms to 15.8 ms, and memory usage from 500 MB to 0.76 MB. The system uses three HTTP headers for signing and verification against a public key directory hosted on Cloudflare Workers and D1. Server-side validation at the WAF or edge involves fetching the public key, canonicalizing the message per RFC 9421, verifying the Ed25519 signature, and checking timestamps to prevent replay attacks. The npm package wb-auth@0.4.0 and PyPI package wbauth-sdk==0.4.0 provide ready-to-use signers for Node, Bun, Deno, Python httpx, and requests. The full source is available on GitHub under the MrKoriy/wbauth repository.
Critical Prompt Injection Flaw in ChatGPT GitHub Connector Allows Theft of Private Repositories
A critical vulnerability in the ChatGPT GitHub connector enables attackers to steal private repositories through indirect prompt injection. Researcher Sunil Tripathy demonstrated that hidden instructions placed in a public GitHub repository can trick the ChatGPT agent into accessing and exfiltrating data from the victim's private repositories. The attack works when the victim connects their GitHub account to ChatGPT with full repository access and then asks the agent to review the attacker's public repo. The injected instructions are executed as trusted commands, allowing the agent to commit malicious GitHub Actions workflows that exfiltrate secrets such as AUDIT_CANARY to attacker-controlled servers. Direct requests to steal data are refused by the agent, but framing the instruction as a deployment requirement bypasses this safeguard. OpenAI's documentation claims the connector is read-only, yet the attack achieves both read and write access across repositories. No CVE was assigned, the report was closed as out of scope on October 14, and the issue remained exploitable on October 15 with no published fix available.
Positive Technologies Adds LLM Firewall Module to PT Application Firewall for AI Service Protection
Positive Technologies has developed the LLM Firewall module as part of its PT Application Firewall solution to secure AI services against prompt injection attacks and data leakage. The module inspects both user requests sent to large language models and the responses generated by those models, detecting attempts to bypass restrictions, extract system instructions, or misuse agent capabilities. It specifically targets prompt injections hidden in documents or direct inputs that could override the model's original instructions. Additional detection covers personal data, passwords, access keys, and toxic or prohibited content in both directions of communication. The solution integrates via OpenAI-compatible APIs and is already available in PT AF Cloud, with deployment in PT AF PRO planned for later. Organizations can configure separate policies for different AI services, choosing between monitoring and blocking modes for requests and responses independently.