Answer: An installed plugin can run code on your machine with your user privileges and can enter Claude's context as instructions. Specifically, it can run hooks and monitors as shell commands, run mods as JavaScript inside Claude Code with your permissions, start MCP and LSP server processes, add its bin/ directory to the Bash tool's PATH, and supply skills, commands, and agents as instructions. Status: Confirmed Entry: tool/claude-code/permissions-and-security/plugin-capabilities Revision: 2 Digest: sha256:5fefbb72999133f4 (changes whenever the content changes) Timeline: Created: 2026-11-13T03:05:26Z Last changed: 2026-11-15T12:24:30Z (Check succeeded again) Last verified: 2026-11-20T12:24:14Z Next check: 2026-11-21T12:24:14Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of General capability | runs code on your machine with your user privileges | [1] | 2026-11-20 MCP servers | A stdio MCP server runs as a process that Claude Code starts on your machine | [1] | 2026-11-20 Skills, commands, and agents | these enter Claude's context as instructions | [1] | 2026-11-20 Permissions scope | run with your full user permissions | [1] | 2026-11-20 Sources: [1] Anthropic. "Plugin security and trust". https://code.claude.com/docs/en/plugins/security.md Role: primary. Accessed: 2026-11-20T12:24:14Z Quote: "A plugin can carry content that runs code on your machine with your user privileges" (Supports: General capability) Quote: "A stdio MCP server runs as a process that Claude Code starts on your machine." (Supports: MCP servers) Quote: "these enter Claude's context as instructions" (Supports: Skills, commands, and agents) Quote: "command hooks and monitors are shell commands that run with your full user permissions" (Supports: Permissions scope) Published: 2026-11-20T03:36:46Z Snapshot SHA-256: c5e86c01a20c93985a5001317e03308bda65d09a6b2fd6adea82487703ff9c65 (captured 2026-11-14T12:19:58Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. On this check the text of [1] was unchanged since 2026-11-15T12:24:30Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T12:24:14Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:24:14Z Last attempt: 2026-11-20T12:24:14Z Notes and limitations: Extracted automatically from a single official source (section "Understand what a plugin can do"). Related: https://latestn.com/tool/claude-code/permissions-and-security/pre-approve-tools-for-skills How do I pre-approve tools for skills in Claude Agent SDK? https://latestn.com/tool/claude-code/permissions-and-security/prompt-injection-protections How does Claude Code protect against prompt injection? https://latestn.com/tool/claude-code/permissions-and-security/relay-permission-prompts How do I relay permission prompts so I can approve or deny tool use remotely in Claude Code? https://latestn.com/tool/claude-code/permissions-and-security/permission-evaluation-order How are permissions evaluated in Claude Code? https://latestn.com/tool/claude-code/permissions-and-security/permission-modes What permission modes does Claude Code support and what does each one do? https://latestn.com/tool/claude-code/permissions-and-security/pin-permission-mode-self-hosted How do I pin a permission mode or pre-approve specific tools for Claude Code self-hosted sessions? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-15T12:24:30Z | recovered | Check succeeded again | — 2026-11-15T12:24:30Z | changed | Skills, commands, and agents: enter Claude's context as instructions → these enter Claude's context as instructions | [1] 2026-11-14T12:20:01Z | check-failed | Extraction did not pass verification: most fields failed verification: fields[0] hooks: quote not found verbatim in the source: "a plugin's hooks run as shell commands at points in Claude Code's lifecycle" | — 2026-11-13T03:05:26Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/claude-code/permissions-and-security/plugin-capabilities/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/claude-code/permissions-and-security/plugin-capabilities POST: https://latestn.com/dispute/tool/claude-code/permissions-and-security/plugin-capabilities with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/claude-code/permissions-and-security/plugin-capabilities?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What can a Claude Code plugin do on my machine? Revision 2, last verified 2026-11-20T12:24:14Z. https://latestn.com/tool/claude-code/permissions-and-security/plugin-capabilities (accessed 2026-11-21T09:53:40Z). @misc{LatestN_tool_claude_code_permissions_and_security_plugin_capabilities_r2, title = {What can a Claude Code plugin do on my machine?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/claude-code/permissions-and-security/plugin-capabilities}}, note = {Revision 2; digest sha256:5fefbb72999133f4; last verified 2026-11-20T12:24:14Z; accessed 2026-11-21T09:53:40Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/claude-code/permissions-and-security/plugin-capabilities