ha-mcp-dev
Not malicious, needs a look Reviewed: expected behaviour
Bawwab's static rules found patterns that deserve a human look. This is not a finding that the package is malicious; the review notes below say what each pattern turned out to be.
Why it was flagged
The static rules marked this “Not malicious, needs a look” because of:
credential-access: Touches credentials or secrets (high, 14× · first at src/ha_mcp/__main__.py:174)remote-code.pipe-shell: Pipes a download into a shell (high, 2× · first at PKG-INFO:189)tool-shadowing: Tool description changes how another tool is used (high, 6× · first at tool ha_config_set_automation:3)
- Kind
- MCP server
- Source
- pypi
- Version
9.3.0.dev3648- Popularity
- 7,515 PyPI downloads/week (2026-11-20)
- Link
- https://pypi.org/project/ha-mcp-dev/9.3.0.dev3648/
- Scanned
- 2026-11-20T22:36:17+00:00
- Scan time
- 40.4 s
Review note: Same as the stable ha-mcp release: install docs, .env config loading, vendored FastMCP/MCP SDK CLI code, and tool descriptions that point to the server's own guide tool.
Findings (23)
remote-code.pipe-shell: Pipes a download into a shell high
PKG-INFO:189
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesOWASP Agentic ASI05 Unexpected Code Execution (RCE)MITRE ATLAS AML.T0050 Command and Scripting InterpreterCWE CWE-494 Download of Code Without Integrity Check
curl -LsSf https://raw.githubusercontent.com/homeassistant-ai/ha-mcp/master/scripts/install-macos.sh | sh
What this means: Executes whatever the remote server returns, unreviewed.
The rule's general advice: Download, verify a checksum, then run. (see the review note for this case)
Expected for what this tool does
Documentation in the package description showing a macOS install script command (from the project's own GitHub repo) for users to run. The package code does not run it.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/__main__.py:174
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
- Or create a .env file in the project directory (copy from .env.example)
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
Help text printed on a missing-configuration error that suggests creating a .env file. No file is read at this line.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/config_settings.py:34
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
env_file = os.getenv("HAMCP_ENV_FILE", ".env")
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Expected for what this tool does
The server loads its own configuration from a .env file in the project directory (name selectable via HAMCP_ENV_FILE) to get its Home Assistant URL and token. The values are used for the user's own Home Assistant connection.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/settings.py:19
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
ENV_FILE = os.getenv("FASTMCP_ENV_FILE", ".env")
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Expected for what this tool does
The vendored FastMCP library reads its own settings from a .env file named by FASTMCP_ENV_FILE. This is standard settings loading.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/client.py:1030
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
console.print(" • ~/.claude.json (Claude Code)")
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
A console message in the vendored FastMCP CLI listing places it searches for MCP servers; it only prints text.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/discovery.py:167
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
"""Scan ``~/.claude.json`` for global and project-scoped MCP servers."""
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Expected for what this tool does
The vendored FastMCP CLI discovery command reads ~/.claude.json to list MCP servers configured for Claude Code. It is a local developer command; the results are printed and not sent anywhere.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/claude_code.py:44
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
Path.home() / ".claude" / "local" / "claude",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
The vendored FastMCP installer checks whether a 'claude' executable exists at ~/.claude/local/claude. It checks the binary path only and reads no credentials.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/claude_desktop.py:173
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
help="Load environment variables from .env file",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
This line is help text or a CLI option description mentioning a .env file; it does not read secrets here. (vendored FastMCP installer CLI)
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/cursor.py:262
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
help="Load environment variables from .env file",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
This line is help text or a CLI option description mentioning a .env file; it does not read secrets here. (vendored FastMCP installer CLI)
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/gemini_cli.py:186
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
help="Load environment variables from .env file",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
This line is help text or a CLI option description mentioning a .env file; it does not read secrets here. (vendored FastMCP installer CLI)
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/goose.py:163
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
help="Load environment variables from .env file",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
This line is help text or a CLI option description mentioning a .env file; it does not read secrets here. (vendored FastMCP installer CLI)
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/mcp_json.py:134
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
help="Load environment variables from .env file",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
This line is help text or a CLI option description mentioning a .env file; it does not read secrets here. (vendored FastMCP installer CLI)
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/fastmcp/cli/install/shared.py:287
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
print(f"[red]Failed to load .env file: {e}[/red]")
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
An error message in the vendored FastMCP installer printed when loading a user-chosen .env file fails. It reads no secrets at this line.
Reviewed by Claude (AI-assisted review), 2026-11-21
remote-code.pipe-shell: Pipes a download into a shell high
src/ha_mcp/_vendor/fastmcp/utilities/mcp_server_config/v1/environments/uv.py:121
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesOWASP Agentic ASI05 Unexpected Code Execution (RCE)MITRE ATLAS AML.T0050 Command and Scripting InterpreterCWE CWE-494 Download of Code Without Integrity Check
"curl -LsSf https://astral.sh/uv/install.sh | sh"
What this means: Executes whatever the remote server returns, unreviewed.
The rule's general advice: Download, verify a checksum, then run. (see the review note for this case)
Expected for what this tool does
An error message in the vendored FastMCP code telling the user how to install uv. It is only text and is not executed.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/_vendor/mcp/cli/cli.py:405
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
help="Load environment variables from a .env file",
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
This line is help text or a CLI option description mentioning a .env file; it does not read secrets here. (vendored MCP SDK CLI)
Reviewed by Claude (AI-assisted review), 2026-11-21
exfil.combo: Reads credentials and makes network calls med
src/ha_mcp/_vendor/mcp/cli/cli.py:280
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0037 Data from Local SystemMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0025 Exfiltration via Cyber Means
env=dict(os.environ.items()), # Copy the environment for subprocess launch
What this means: Credential access together with outbound requests in one file is the shape of data theft.
The rule's general advice: Do not install unless the flow is understood and benign. (see the review note for this case)
Known false alarm
The environment copy is passed only to the MCP Inspector (npx) subprocess started by the vendored MCP SDK 'dev' command. It is not sent over the network.
Reviewed by Claude (AI-assisted review), 2026-11-21
credential-access: Touches credentials or secrets high
src/ha_mcp/settings_ui/_handlers_backups.py:454
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
Standalone (file) mode only: an env var (process or ``.env``) wins over
What this means: Reads files or environment holding keys, tokens or sessions.
The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)
Known false alarm
A docstring explaining that an environment variable (from the process or .env) overrides the settings file, so a UI edit to that field is refused. No credential is read at this line.
Reviewed by Claude (AI-assisted review), 2026-11-21
tool-shadowing: Tool description changes how another tool is used high
tool ha_config_set_automation:3
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
MUST call ha_get_skill_guide
What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.
The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)
Expected for what this tool does
These tool descriptions tell the model to call the server's own 'ha_get_skill_guide' tool (or consult local skills) before writing Home Assistant configuration. It refers to a tool in the same server and does not redirect other tools.
Reviewed by Claude (AI-assisted review), 2026-11-21
tool-shadowing: Tool description changes how another tool is used high
tool ha_config_set_dashboard:3
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
MUST call ha_get_skill_guide
What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.
The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)
Expected for what this tool does
These tool descriptions tell the model to call the server's own 'ha_get_skill_guide' tool (or consult local skills) before writing Home Assistant configuration. It refers to a tool in the same server and does not redirect other tools.
Reviewed by Claude (AI-assisted review), 2026-11-21
tool-shadowing: Tool description changes how another tool is used high
tool ha_config_set_helper:4
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
MUST call ha_get_skill_guide
What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.
The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)
Expected for what this tool does
These tool descriptions tell the model to call the server's own 'ha_get_skill_guide' tool (or consult local skills) before writing Home Assistant configuration. It refers to a tool in the same server and does not redirect other tools.
Reviewed by Claude (AI-assisted review), 2026-11-21
tool-shadowing: Tool description changes how another tool is used high
tool ha_config_set_scene:3
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
MUST call ha_get_skill_guide
What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.
The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)
Expected for what this tool does
These tool descriptions tell the model to call the server's own 'ha_get_skill_guide' tool (or consult local skills) before writing Home Assistant configuration. It refers to a tool in the same server and does not redirect other tools.
Reviewed by Claude (AI-assisted review), 2026-11-21
tool-shadowing: Tool description changes how another tool is used high
tool ha_config_set_script:3
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
MUST call ha_get_skill_guide
What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.
The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)
Expected for what this tool does
These tool descriptions tell the model to call the server's own 'ha_get_skill_guide' tool (or consult local skills) before writing Home Assistant configuration. It refers to a tool in the same server and does not redirect other tools.
Reviewed by Claude (AI-assisted review), 2026-11-21
tool-shadowing: Tool description changes how another tool is used high
tool ha_config_set_yaml:3
OWASP LLM LLM01 Prompt InjectionOWASP Agentic ASI02 Tool Misuse and ExploitationOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0110.000 AI Agent Tool Poisoning: Definition and InstructionsMITRE ATLAS AML.T0011.002 User Execution: Poisoned AI Agent ToolCWE CWE-1427 Improper Neutralization of Input Used for LLM Prompting
MUST call ha_get_skill_guide
What this means: A description that names a different tool and tells the AI how to use it (or to use it instead) can quietly hijack that tool, for example by adding a recipient to emails.
The rule's general advice: Do not install; a tool should only describe itself. (see the review note for this case)
Expected for what this tool does
These tool descriptions tell the model to call the server's own 'ha_get_skill_guide' tool (or consult local skills) before writing Home Assistant configuration. It refers to a tool in the same server and does not redirect other tools.
Reviewed by Claude (AI-assisted review), 2026-11-21
9 low-severity notes (informational, do not affect the result)
credential-access: Touches credentials or secrets low
src/ha_mcp/config_backup.py:32
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
(includes values loaded from ``.env`` via ``load_dotenv`` at
What this means: Reads files or environment holding keys, tokens or sessions.
What you can do: Confirm the tool needs this; otherwise do not install.
credential-access: Touches credentials or secrets low
src/ha_mcp/config_overrides.py:46
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
(includes values loaded from ``.env`` via ``load_dotenv`` at
What this means: Reads files or environment holding keys, tokens or sessions.
What you can do: Confirm the tool needs this; otherwise do not install.
credential-access: Touches credentials or secrets low
src/ha_mcp/_vendor/fastmcp/cli/cli.py:307
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
env = dict(os.environ.items()) | env_vars | {"FASTMCP_UV_SPAWNED": "1"}
What this means: Reads files or environment holding keys, tokens or sessions.
What you can do: Confirm the tool needs this; otherwise do not install.
credential-access: Touches credentials or secrets low
src/ha_mcp/_vendor/fastmcp/client/transports/stdio.py:518
OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration
env = os.environ.copy()
What this means: Reads files or environment holding keys, tokens or sessions.
What you can do: Confirm the tool needs this; otherwise do not install.
install-dir-access: Reads or writes Claude's install folders low
src/ha_mcp/_vendor/fastmcp/server/providers/skills/__init__.py:19
MITRE ATLAS AML.T0081 Modify AI Agent Configuration
mcp.add_provider(SkillProvider(Path.home() / ".claude/skills/pdf-processing"))
What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.
What you can do: Check which files it adds there and that you want them.
install-dir-access: Reads or writes Claude's install folders low
src/ha_mcp/_vendor/fastmcp/server/providers/skills/claude_provider.py:12
MITRE ATLAS AML.T0081 Modify AI Agent Configuration
"""Provider for Claude Code skills from ~/.claude/skills/.
What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.
What you can do: Check which files it adds there and that you want them.
install-dir-access: Reads or writes Claude's install folders low
src/ha_mcp/_vendor/fastmcp/server/providers/skills/directory_provider.py:65
MITRE ATLAS AML.T0081 Modify AI Agent Configuration
roots=Path.home() / ".claude" / "skills",
What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.
What you can do: Check which files it adds there and that you want them.
install-dir-access: Reads or writes Claude's install folders low
src/ha_mcp/_vendor/fastmcp/server/providers/skills/skill_provider.py:219
MITRE ATLAS AML.T0081 Modify AI Agent Configuration
Path.home() / ".claude/skills/pdf-processing"
What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.
What you can do: Check which files it adds there and that you want them.
install-dir-access: Reads or writes Claude's install folders low
src/ha_mcp/_vendor/fastmcp/utilities/skills.py:161
MITRE ATLAS AML.T0081 Modify AI Agent Configuration
"~/.claude/skills"
What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.
What you can do: Check which files it adds there and that you want them.
Dispute this result
Maintainer and think this is wrong? Open a dispute on GitHub with the target id mcp-pypi-ha-mcp-dev-20e0cd.