---
id: "CVE-2026-62331"
kind: "vulnerability"
aliases: ["CVE-2026-62331"]
title: "praktor: Telegram file upload lets remote users write outside uploads folder"
severity: "high"
cvss: 8.2
epss: null
exploited: false
status: "open"
category: "ai-tooling"
products: []
ecosystems: []
tags: ["ai-tooling", "claude-code", "docker", "praktor", "telegram"]
not_affected_if: ["allow-from-restricted", "telegram-disabled", "version-0.9.1-or-later"]
published_at: "2026-10-30T19:18:33+00:00"
first_seen_at: "2026-10-30T19:25:30+00:00"
updated_at: "2026-10-30T19:25:54+00:00"
sources: [{"source": "nvd", "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-62331", "first_seen_at": "2026-10-30T19:25:30+00:00"}]
---

# praktor: Telegram file upload lets remote users write outside uploads folder

praktor is a multi-agent Claude Code orchestrator driven through Telegram. Its Telegram upload handler joins the attacker-supplied filename into a path without sanitizing it, so a remote Telegram user can write files into the agent workspace and plant instruction files like CLAUDE.md that later agent runs read and act on.

_Summary, check and fix written by a model from the sources; labelled as such._

## How to check

- Check the praktor version in use; affected if below `0.9.1`
- Check `allow_from` in the praktor configuration; `allow_from: []` lets any Telegram sender reach the upload path
- Look for unexpected `CLAUDE.md`, skill, hook or config files under the `/workspace/agent` volume

## Fix

- praktor: move to `0.9.1`

## Timeline

- 2026-10-30T19:25:30+00:00 nvd https://nvd.nist.gov/vuln/detail/CVE-2026-62331

## From the sources

praktor is a multi-agent Claude Code orchestrator with Telegram input and output, Docker isolation, and a Mission Control UI. Prior to 0.9.1, both Telegram upload sites in `internal/telegram/bot.go` concatenate the attacker-controlled document filename into a destination path without sanitization, and `path.Join` collapses parent-directory traversal. A remote Telegram user who can message the bot can therefore write outside `uploads/` within the persistent agent workspace mounted at `/workspace/agent`. Because `allow_from: []` permits all senders by default, no prior authentication or privilege is required in that configuration. The attacker can plant or overwrite `CLAUDE.md`, instruction, skill, hook, or configuration files that the Claude agent reads and operates on across later runs, turning a document upload into stored instruction injection. The demonstrated arbitrary-write scope is the persistent agent workspace volume, not an unrestricted host-filesystem write. This issue is fixed in version 0.9.1.

---

_Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai_