PraisonAI Dynamic Context history and terminal tools read files outside configured storage via path traversal
CVE-2026-56833 · GHSA-22cj-m4wf-fv2c · published 2026-06-18 · updated 2026-07-20
- Severity
- High
- CVSS score
- 7.5
- Chance of exploitation (EPSS)
- 0.46%
- Fixed in
- 4.6.59
- Exploitation No known exploitation
- Automatable Yes
- Technical impact Partial
- CISA SSVC
If a PraisonAI application exposes these Dynamic Context tools to untrusted or lower-trust prompts, the lower-trust caller can read files outside the configured context storage when the target file can be reached with the tool-imposed suffix: history_* tools can disclose reachable .jsonl files; terminal_* tools can disclose reachable .log files; and cross-run or cross-agent context/history/logs can be disclosed if their path is known or guessable. This can expose conversation history, prompts, terminal output, command logs, tokens, API keys, cloud credentials, operational data, or other secrets stored in JSONL/log files readable by the PraisonAI process. …
Upgrade praisonai to 4.6.59 or later. If you start it from an MCP config with npx or uvx, pin that version so an old cached copy is not used.
Treat run_id and agent_id as logical identifiers, not path components. Recommended fixes: 1. Reject absolute paths, path separators, and traversal components in run_id and agent_id. 2. Build candidate paths, call .resolve(), and reject any path that is not under self.base_dir.resolve(). 3. Apply the same containment helper to history append/read/search/clear/export and terminal log/read/search/clear/export paths. 4. Prefer opaque server-generated run and agent IDs in tool schemas. 5. Pair this with an identifier allowlist, because run_id and agent_id should not need filesystem syntax.
| Package | Registry | Affected | Fixed in |
|---|---|---|---|
| praisonai | PyPI | >= 3.8.1, <= 4.6.58 | 4.6.59 |
- CriticalMCP Path Traversal Allows Full-Filesystem Read and JWT Secret Exfiltration Leading to Account Takeoverlangflow, langflow-base
- Critical@zereight/mcp-gitlab: Unauthenticated arbitrary file read via `upload_markdown` enables PAT exfiltration and full account takeover@zereight/mcp-gitlab
- CriticalFlowise RCE via SQL Database Chain Nodeflowise, flowise-components
- CriticalLangroid: SQLChatAgent dangerous-function blocklist can be bypassed with quoted or schema-qualified pg_read_file callslangroid
- Criticalfast-mcp-telegram: Bearer token path traversal bypasses reserved Telegram session protectionfast-mcp-telegram
- CriticalMCP Toolbox HTTP tool path parameters allow traversal to unintended endpointsgithub.com/googleapis/mcp-toolbox
- CriticalPraisonAI: AICoder Arbitrary File Write and Command Execution via LLM Tool Callspraisonai
- CriticalPraisonAI: Code Injection via f-string Interpolation in Deploy API Server Generationpraisonai
- CriticalPraisonAI has a Browser Server WebSocket origin validation bypass via unanchored regex (patch bypass of CVE-2026-40289 / GHSA-8x8f-54wf-vv92praisonai
- CriticalPraisonAI: Arbitrary File Read/Write via `multiedit` Tool Without Path Validationpraisonai
- CriticalPraisonAI: Unauthenticated RCE via Jobs API + Approval Bypasspraisonaiagents, praisonai
- CriticalPraisonAI: AgentOS remains unauthenticated after incomplete fix version and allows remote agent invocationpraisonai
The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.