NousResearch/hermes-agent skill
agent framework
Hermes Agent is a self-hosted AI agent server that creates and improves its own skills from conversations, maintains persistent memory across sessions, and connects to messaging platforms (Telegram, Discord, Slack, WhatsApp, Signal) and CLI interfaces. It supports multiple LLM providers and includes a built-in scheduler for automations, parallel subagent execution, and integration with various compute backends from local machines to serverless infrastructure.
Run an AI agent on any infrastructure—laptop, $5 VPS, GPU cluster, or serverless (Modal, Daytona) Talk to your agent from CLI, Telegram, Discord, Slack, or other messaging platforms through a single gateway Delegate complex tasks to spawned subagents and parallelize workstreams without losing context Schedule automations like daily reports or nightly backups to run unattended in natural language Let the agent autonomously build and refine skills from past conversations using its learning loop 3 CRITICAL✓ 80 HIGH✓ 563 MEDIUM 499 LOW 5 INFO
✓ CRITICAL/HIGH reflect AI-verified findings (false positives excluded) · MEDIUM/LOW/INFO are unverified scanner output
AI-verified (CRITICAL/HIGH): 3 confirmed (4%) 66 likely real (96%) 131 false positive — excluded from CRITICAL/HIGH count above
Findings by checker · 8 high-signal, 12 mostly false-positive (hidden by default)
CHK-083 25 findings 19 likely 6 false positive
24% FP CHK-047 18 findings 16 likely 2 false positive
11% FP CHK-099 6 findings 6 likely
0% FP CHK-081 3 findings 3 likely
0% FP CHK-074 1 finding 1 likely
0% FP CHK-090 1 finding 1 likely
0% FP CHK-103 1 finding 1 likely
0% FP CHK-105 1 finding 1 likely
0% FP ▼ Show 12 checkers that are mostly false positives (144 findings) 69 findings click to expand
CHK-081 Command injection risk — exec/execSync with string interpolation in tools/transcription_tools.py
tools/transcription_tools.py
AI: likely real confirmed ▼
CHK-081 Command injection risk — exec/execSync with string interpolation in plugins/security-guidance/patterns.py
plugins/security-guidance/patterns.py
AI: likely real confirmed ▼
CHK-081 Command injection risk — exec/execSync with string interpolation in apps/desktop/electron/backend-probes.ts
apps/desktop/electron/backend-probes.ts
AI: likely real confirmed ▼
CHK-129 Sensitive field in return type/schema — token: str
hermes_cli/telegram_managed_bot.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in hermes_cli/gateway.py
hermes_cli/gateway.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/dashboard_auth/nous/__init__.py
plugins/dashboard_auth/nous/__init__.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/spotify/client.py
plugins/spotify/client.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/platforms/photon/auth.py
plugins/platforms/photon/auth.py
AI: likely real likely ▼
CHK-074 Permission bypass referenced in documentation
optional-skills/autonomous-ai-agents/antigravity-cli/SKILL.md
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/web/searxng/provider.py
plugins/web/searxng/provider.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in scripts/ci/timings_report.py
scripts/ci/timings_report.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in hermes_cli/auth.py
hermes_cli/auth.py
AI: likely real likely ▼
CHK-115 Skill instructs agent to access credential file: .env`
website/docs/user-guide/skills/optional/security/security-1password.md
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in agent/pet/store.py
agent/pet/store.py
AI: likely real likely ▼
CHK-129 Sensitive field in return type/schema — api_key: str
hermes_cli/model_switch.py
AI: likely real likely ▼
CHK-023 Embedded instruction in skill file — directive to model
optional-skills/security/godmode/SKILL.md
AI: confirmed likely ▼
CHK-114 Cyrillic homoglyph substitution detected in skill file
optional-skills/security/godmode/SKILL.md
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/web/xai/provider.py
plugins/web/xai/provider.py
AI: likely real likely ▼
CHK-115 Skill instructs agent to access credential file: .env`
website/docs/user-guide/skills/optional/productivity/productivity-canvas.md
AI: likely real likely ▼
CHK-090 HTTP transport bound to 0.0.0.0 — exposed to all network interfaces
hermes_cli/webhook.py
AI: likely real likely ▼
CHK-029 Dynamic MCP server instantiation in skill
optional-skills/mcp/mcporter/SKILL.md
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/web/tavily/provider.py
plugins/web/tavily/provider.py
AI: likely real likely ▼
CHK-125 subprocess module — arbitrary shell execution — no scope constraint in scripts/gws_bridge.py
skills/productivity/google-workspace/scripts/gws_bridge.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/memory/openviking/__init__.py
plugins/memory/openviking/__init__.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: optional-skills/blockchain/hyperliquid/scripts/hyperliquid_client.py
optional-skills/blockchain/hyperliquid/scripts/hyperliquid_client.py
AI: likely real possible ▼
CHK-129 Sensitive field in return type/schema — secret: str
tools/microsoft_graph_auth.py
AI: likely real likely ▼
CHK-129 Sensitive field in return type/schema — token: str
tools/managed_tool_gateway.py
AI: likely real likely ▼
CHK-125 exec() — arbitrary code execution — no scope constraint in references/refusal-detection.md
optional-skills/security/godmode/references/refusal-detection.md
AI: likely real likely ▼
CHK-099 Potential IDOR — 'task_id' accessed without ownership check
tools/browser_dialog_tool.py
AI: likely real possible ▼
CHK-099 Potential IDOR — 'task_id' accessed without ownership check
tools/vision_tools.py
AI: likely real possible ▼
CHK-099 Potential IDOR — 'thread_id' accessed without ownership check
tools/cronjob_tools.py
AI: likely real possible ▼
CHK-083 Potential path traversal — user-controlled path in file operation: agent/file_safety.py
agent/file_safety.py
AI: likely real possible ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/dashboard_auth/self_hosted/__init__.py
plugins/dashboard_auth/self_hosted/__init__.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: tools/code_execution_tool.py
tools/code_execution_tool.py
AI: likely real possible ▼
CHK-099 Potential IDOR — 'task_id' accessed without ownership check
tools/file_tools.py
AI: likely real likely ▼
CHK-125 exec() — arbitrary code execution — no scope constraint in scripts/godmode_race.py
optional-skills/security/godmode/scripts/godmode_race.py
AI: confirmed likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: tools/voice_mode.py
tools/voice_mode.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: hermes_cli/web_git.py
hermes_cli/web_git.py
AI: likely real possible ▼
CHK-083 Potential path traversal — user-controlled path in file operation: optional-skills/security/godmode/scripts/godmode_race.py
optional-skills/security/godmode/scripts/godmode_race.py
AI: likely real possible ▼
CHK-083 Potential path traversal — user-controlled path in file operation: hermes_cli/browser_connect.py
hermes_cli/browser_connect.py
AI: likely real likely ▼
CHK-099 Potential IDOR — 'user_id' accessed without ownership check
tools/process_registry.py
AI: likely real possible ▼
CHK-103 Destructive operation without authorization check
tools/process_registry.py
AI: likely real possible ▼
CHK-099 Potential IDOR — 'task_id' accessed without ownership check
tools/browser_cdp_tool.py
AI: likely real possible ▼
CHK-125 subprocess module — arbitrary shell execution — no scope constraint in hermes_cli/cli_commands_mixin.py
hermes_cli/cli_commands_mixin.py
AI: likely real likely ▼
CHK-125 exec() — arbitrary code execution — no scope constraint in environments/daytona.py
tools/environments/daytona.py
AI: likely real likely ▼
CHK-125 subprocess module — arbitrary shell execution — no scope constraint in security-guidance/patterns.py
plugins/security-guidance/patterns.py
AI: confirmed likely ▼
CHK-125 exec() — arbitrary code execution — no scope constraint in godmode/SKILL.md
optional-skills/security/godmode/SKILL.md
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: agent/transports/codex_app_server.py
agent/transports/codex_app_server.py
AI: likely real possible ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in hermes_cli/telegram_managed_bot.py
hermes_cli/telegram_managed_bot.py
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in tools/skills_hub.py
tools/skills_hub.py
AI: likely real likely ▼
CHK-125 eval() — arbitrary code execution — no scope constraint in scripts/profile-synth-stream.mjs
apps/desktop/scripts/profile-synth-stream.mjs
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: optional-skills/security/godmode/scripts/load_godmode.py
optional-skills/security/godmode/scripts/load_godmode.py
AI: likely real possible ▼
CHK-083 Potential path traversal — user-controlled path in file operation: scripts/ci/timings_report.py
scripts/ci/timings_report.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: tools/browser_tool.py
tools/browser_tool.py
AI: likely real likely ▼
CHK-115 Skill instructs agent to access credential file: .env`
optional-skills/security/1password/SKILL.md
AI: likely real likely ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in hermes_cli/doctor.py
hermes_cli/doctor.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: scripts/docker_rebootstrap_nous_session.py
scripts/docker_rebootstrap_nous_session.py
AI: likely real possible ▼
CHK-115 Skill instructs agent to access credential file: .env`
optional-skills/productivity/canvas/SKILL.md
AI: likely real likely ▼
CHK-105 Secret echoed to CI logs in .github/workflows/tests.yml
.github/workflows/tests.yml
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: optional-skills/security/godmode/scripts/parseltongue.py
optional-skills/security/godmode/scripts/parseltongue.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: agent/lsp/servers.py
agent/lsp/servers.py
AI: likely real possible ▼
CHK-083 Potential path traversal — user-controlled path in file operation: hermes_cli/kanban_db.py
hermes_cli/kanban_db.py
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: optional-skills/productivity/canvas/scripts/canvas_api.py
optional-skills/productivity/canvas/scripts/canvas_api.py
AI: likely real possible ▼
CHK-115 Skill instructs agent to access credential file: .env`
website/docs/user-guide/skills/bundled/productivity/productivity-airtable.md
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: hermes_cli/urllib_security.py
hermes_cli/urllib_security.py
AI: likely real possible ▼
CHK-047 Potential SSRF — user-controlled URL passed to HTTP client in plugins/memory/honcho/oauth.py
plugins/memory/honcho/oauth.py
AI: likely real likely ▼
CHK-023 Embedded instruction in skill file — directive to model
website/docs/user-guide/skills/optional/security/security-godmode.md
AI: likely real likely ▼
CHK-083 Potential path traversal — user-controlled path in file operation: hermes_cli/stdio.py
hermes_cli/stdio.py
AI: likely real possible ▼
CHK-083 Potential path traversal — user-controlled path in file operation: tools/environments/local.py
tools/environments/local.py
AI: likely real likely ▼
▼ Show 131 false positives (66% of this view) Last scanned: Jul 11, 2026
More servers
google-gemini/gemini-cli 85
An open-source AI agent that brings the power of Gemini directly into your terminal.
106k★
skypilot-org/skypilot 85
obra/superpowers 85
Foundational skill pack by Jesse Vincent now in anthropics/claude-plugins-official. Includes ffuf web-fuzzing/pentest skill. Partial analysis done — full hook and plugin inspection pending. tier=T2
191k★
Significant-Gravitas/AutoGPT 85
AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.
186k★
nanocoai/nanoclaw 85
A lightweight alternative to OpenClaw that runs in containers for security. Connects to WhatsApp, Telegram, Slack, Discord, Gmail and other messaging apps,, has memory, scheduled jobs, and runs directly on Anthropic's Agents SDK
30k★
google-gemini/gemini-cli 85
An open-source AI agent that brings the power of Gemini directly into your terminal.
106k★
Scan your entire org's MCP deployment
2,500+ repos pre-scored. 22% carry CRITICAL findings.