Skip to main content
guidesSource-backed

Plugins in Claude Agent SDK Deployments

A practical walkthrough of loading local plugins in the Claude Agent SDK: the plugins option with type local and a path, what plugins bundle (skills, agents, hooks, MCP servers), verifying loads via the init message, and invoking namespaced plugin skills.

by JPette1783·added 2026-06-05·
Review first review before installing

Open the source and read safety notes before installing.

Citation facts

Source-backed facts for citing this resource, derived directly from the registry — also available as plain text for AI assistants.

Source URLs
https://code.claude.com/docs/en/agent-sdk/plugins, https://github.com/JSONbored/awesome-claude/blob/main/content/guides/plugins-in-claude-agent-sdk-deployments.mdx
Safety notes
The SDK only accepts type: local plugins; to use a marketplace plugin, download it first and point at the local directory., Plugins can bundle hooks, MCP servers, and agents that run with your application's privileges; review a plugin's contents before loading it., Check the system init message's plugin_errors to fail fast in CI when a plugin or a dependency version did not load.
Privacy notes
Bundled MCP servers and hooks can access local files and external services; confirm what a plugin's components reach before deploying it., Plugin skills and commands are sent to the model as available capabilities; keep secrets out of plugin definitions., Loading a plugin from a path you do not control is a supply-chain decision; prefer plugins you have reviewed and pinned.
Author
JPette1783
Submitted by
JPette1783
Claim status
unclaimed
Last verified
2026-06-05

Decision playbook

Review trust signals before you adopt

Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.

Compare context
Selected

0

Current score

63

Baseline

Delta

No baseline selected

No major trust-signal divergence detected in the current selection.

Source and provenance checks

Needs review

Confirm ownership and provenance before trusting install instructions.

  • Source link availableRequired

    Open the canonical repository and verify ownership.

    Done
  • Source provenance statusRequired

    Marked as source-backed.

    Done
  • Metadata reviewed

    No reviewed flag detected in metadata.

    Pending

Safety and privacy checks

Complete

Validate risk disclosures before installation or API wiring.

  • Safety notes presentRequired

    Review the listed safety guidance before running commands.

    Done
  • Privacy notes presentRequired

    Review data handling notes before connecting accounts or secrets.

    Done
  • Trust level risk gateRequired

    Trust level does not block evaluation.

    Done

Package and install checks

Needs review

Check package metadata and artifact integrity signals.

  • Install payload available

    Install or copy payload is available for review.

    Done
  • Package verification flag

    No package verification flag provided.

    Pending
  • Checksum metadata

    No checksum provided for downloaded artifact.

    Pending

Compare-driven decision checks

Needs review

Use compare context to validate trade-offs before adoption.

  • Compare tray has multiple entries

    Add at least one more entry to compare trust differences.

    Pending
  • Baseline comparison available

    No baseline peer selected yet.

    Pending
  • Diverging trust signals identified

    No major trust-signal divergence found.

    Pending

Setup at a glance

Copy & paste

Copy-ready — paste the snippet to get started.

Adoption plan

Balanced adoption plan

Current risk score 24/100. Use staged verification before broader rollout.

Risk 24

Pre-adoption checks

Validate source and review signals before any execution.

  • Confirm source provenanceRequired

    Source URL/provenance metadata is present.

    Done
  • Confirm metadata review state

    No review metadata found; increase manual validation.

    Pending
  • Verify install payload

    Install/config payload exists and can be inspected.

    Done

Security checks

Confirm safety, privacy, and package integrity signals.

  • Review safety notesRequired

    Safety notes are present.

    Done
  • Review privacy notesRequired

    Privacy notes are present.

    Done
  • Verify package integrity metadata

    No package verification/checksum metadata.

    Pending

Rollout

Adopt in controlled steps based on the selected plan.

  • Run in isolated sandbox firstRequired

    Use a constrained sandbox and observe behavior across multiple tasks.

    Pending
  • Roll out graduallyRequired

    Roll out to a small cohort before wider usage.

    Pending
  • Set monitoring and fallback

    Define rollback path and monitor errors after adoption.

    Pending

Evidence readiness

Evidence readiness matrix · balanced

Missing required evidence: Metadata review. Risk score 31.

Risk 31

Source provenance

Present

Source repository/provenance is listed.

Required in this preset

Metadata review

Missing

Review metadata is missing.

Required in this preset

Safety notes

Present

Safety notes are present.

Required in this preset

Privacy notes

Present

Privacy notes are present.

Optional in this preset

Package integrity

Missing

Package integrity metadata is missing.

Optional in this preset

Install payload

Present

Install payload is available.

Required in this preset

Required gaps: Metadata review

Decision timeline

Decision timeline · balanced

Blocking gaps: Check metadata review status. Risk 28.

Risk 28

triage

Confirm source provenanceRequired

Source/provenance metadata is available.

Done

triage

Check metadata review statusRequired

Review metadata is missing.

Pending

verify

Review safety notesRequired

Safety notes are available.

Done

verify

Review privacy notes

Privacy notes are available.

Done

verify

Validate package integrity metadata

Package integrity metadata is missing.

Pending

rollout

Verify install payload and commandsRequired

Install payload is available.

Done

Blockers: Check metadata review status

Prerequisite readiness

Prerequisite readiness

3 prerequisites to line up before setup.

0/3 ready
Install & runtime2General1

Safety & privacy surface

Safety & privacy surface

3 safety and 3 privacy notes across 6 risk areas. Review closely: credentials & tokens, network access, third-party handling.

6 areas
  • SafetyNetwork accessThe SDK only accepts type: local plugins; to use a marketplace plugin, download it first and point at the local directory.
  • SafetyExecution & processesPlugins can bundle hooks, MCP servers, and agents that run with your application's privileges; review a plugin's contents before loading it.
  • SafetyGeneralCheck the system init message's plugin_errors to fail fast in CI when a plugin or a dependency version did not load.
  • PrivacyThird-party handlingBundled MCP servers and hooks can access local files and external services; confirm what a plugin's components reach before deploying it.
  • PrivacyCredentials & tokensPlugin skills and commands are sent to the model as available capabilities; keep secrets out of plugin definitions.
  • PrivacyLocal filesLoading a plugin from a path you do not control is a supply-chain decision; prefer plugins you have reviewed and pinned.

Safety notes

  • The SDK only accepts type: local plugins; to use a marketplace plugin, download it first and point at the local directory.
  • Plugins can bundle hooks, MCP servers, and agents that run with your application's privileges; review a plugin's contents before loading it.
  • Check the system init message's plugin_errors to fail fast in CI when a plugin or a dependency version did not load.

Privacy notes

  • Bundled MCP servers and hooks can access local files and external services; confirm what a plugin's components reach before deploying it.
  • Plugin skills and commands are sent to the model as available capabilities; keep secrets out of plugin definitions.
  • Loading a plugin from a path you do not control is a supply-chain decision; prefer plugins you have reviewed and pinned.

Prerequisites

  • The Claude Agent SDK installed for Python or TypeScript.
  • One or more plugin directories on the local filesystem (the parent of skills/, agents/, hooks/, or .claude-plugin/).
  • For marketplace plugins, the plugin downloaded locally so you can provide a path.

Schema details

Install type
copy
Troubleshooting
Yes
Full copyable content
## Overview

Plugins package Claude Code extensions (skills, agents, hooks, MCP servers) into
a shareable unit. The Claude Agent SDK can load plugins from local directories so
your agent sessions gain those capabilities programmatically.

## Load plugins

Provide local paths via the `plugins` option. The `type` must be `"local"` (the
only value the SDK accepts); for a marketplace plugin, download it first and pass
the local path.

```typescript
for await (const message of query({
  prompt: "Hello",
  options: {
    plugins: [
      { type: "local", path: "./my-plugin" },
      { type: "local", path: "/absolute/path/to/another-plugin" },
    ],
  },
})) { /* ... */ }
```

The path should point at the plugin root (the parent of `skills/`, `agents/`,
`hooks/`, or `.claude-plugin/`), not a subdirectory. Relative paths resolve from
the current working directory.

## Verify the load

When plugins load, they appear in the `system` `init` message. Check it (and its
`plugin_errors`) to confirm loads and fail CI on errors:

```typescript
if (message.type === "system" && message.subtype === "init") {
  console.log("Plugins:", message.plugins);          // [{ name, path }]
  console.log("Skills:", message.skills);            // ["my-plugin:greet"]
  console.log("Commands:", message.slash_commands);
}
```

## Invoke plugin skills

Plugin skills are namespaced with the plugin name to avoid conflicts. Invoke one
directly by sending `/plugin-name:skill-name` as the prompt:

```typescript
for await (const message of query({
  prompt: "/my-plugin:greet",
  options: { plugins: [{ type: "local", path: "./my-plugin" }] },
})) { /* ... */ }
```

CLI-installed plugins live under `~/.claude/plugins/`; point the SDK at that path
to reuse them.

## Common uses

- Load a plugin during development without installing it globally.
- Include project plugins in your repo for team-wide consistency.
- Combine plugins from multiple local locations in one session.

## Troubleshooting

- If a plugin does not appear, check the path points at the plugin root, validate
  any `plugin.json`, and confirm the directory is readable.
- If skills do not work, invoke them with the `/plugin-name:skill-name` namespace
  and confirm they appear in the init message's `skills` list.

## Source

- Plugins in the SDK: https://code.claude.com/docs/en/agent-sdk/plugins

About this resource

Overview

Plugins package Claude Code extensions (skills, agents, hooks, MCP servers) into a shareable unit. The Claude Agent SDK can load plugins from local directories so your agent sessions gain those capabilities programmatically.

Load plugins

Provide local paths via the plugins option. The type must be "local" (the only value the SDK accepts); for a marketplace plugin, download it first and pass the local path.

for await (const message of query({
  prompt: "Hello",
  options: {
    plugins: [
      { type: "local", path: "./my-plugin" },
      { type: "local", path: "/absolute/path/to/another-plugin" },
    ],
  },
})) { /* ... */ }

The path should point at the plugin root (the parent of skills/, agents/, hooks/, or .claude-plugin/), not a subdirectory. Relative paths resolve from the current working directory.

Verify the load

When plugins load, they appear in the system init message. Check it (and its plugin_errors) to confirm loads and fail CI on errors:

if (message.type === "system" && message.subtype === "init") {
  console.log("Plugins:", message.plugins);          // [{ name, path }]
  console.log("Skills:", message.skills);            // ["my-plugin:greet"]
  console.log("Commands:", message.slash_commands);
}

Invoke plugin skills

Plugin skills are namespaced with the plugin name to avoid conflicts. Invoke one directly by sending /plugin-name:skill-name as the prompt:

for await (const message of query({
  prompt: "/my-plugin:greet",
  options: { plugins: [{ type: "local", path: "./my-plugin" }] },
})) { /* ... */ }

CLI-installed plugins live under ~/.claude/plugins/; point the SDK at that path to reuse them.

Common uses

  • Load a plugin during development without installing it globally.
  • Include project plugins in your repo for team-wide consistency.
  • Combine plugins from multiple local locations in one session.

Troubleshooting

  • If a plugin does not appear, check the path points at the plugin root, validate any plugin.json, and confirm the directory is readable.
  • If skills do not work, invoke them with the /plugin-name:skill-name namespace and confirm they appear in the init message's skills list.

Source

Source citations

Add this badge to your README

Show that Plugins in Claude Agent SDK Deployments is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.

Listed on HeyClaude
[![Listed on HeyClaude](https://heyclau.de/badge/guides/plugins-in-claude-agent-sdk-deployments.svg)](https://heyclau.de/entry/guides/plugins-in-claude-agent-sdk-deployments)

How it compares

Plugins in Claude Agent SDK Deployments side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.

2 trust signals differ across this comparison (Source provenance, Submitter).

Field

A practical walkthrough of loading local plugins in the Claude Agent SDK: the plugins option with type local and a path, what plugins bundle (skills, agents, hooks, MCP servers), verifying loads via the init message, and invoking namespaced plugin skills.

Open dossier

A practical walkthrough of using Agent Skills in the Claude Agent SDK: how skills are discovered from the filesystem via settingSources, the skills option to enable or filter them, tool access, and troubleshooting discovery.

Open dossier

A practical walkthrough of mirroring Claude Agent SDK session transcripts to external storage: the SessionStore interface, the SessionKey shape, resuming across hosts, reference adapters for S3/Redis/Postgres, and behavior notes.

Open dossier

Production-oriented quickstart for the Claude Agent SDK: project bootstrap with agent-sdk-dev plugin, query loops, permissions, observability hooks, and deployment guardrails.

Open dossier
Next steps
Trust
Review statusNot reviewedNot reviewedNot reviewedNot reviewed
Package trustPackage not verifiedPackage not verifiedPackage not verifiedPackage not verified
Source provenanceDiffersSource-backedSource-backedSource-backedSubmission linkedSource submission
SubmitterDiffersJPette1783JPette1783JPette1783kiannidev
Install riskReview firstReview firstReview firstReview first
Notes Safety ✓ Privacy ✓ Safety ✓ Privacy ✓ Safety ✓ Privacy ✓ Safety ✓ Privacy ✓
Brand
Categoryguidesguidesguidesguides
SourceSource-backedSource-backedSource-backedSource-backed
AuthorJPette1783JPette1783JPette1783kiannidev
Added2026-06-052026-06-052026-06-052026-06-14
Platforms
Harness
Source repo
Safety notesThe SDK only accepts type: local plugins; to use a marketplace plugin, download it first and point at the local directory. Plugins can bundle hooks, MCP servers, and agents that run with your application's privileges; review a plugin's contents before loading it. Check the system init message's plugin_errors to fail fast in CI when a plugin or a dependency version did not load.The skills option is a context filter, not a sandbox: unlisted skills are hidden from the model but their files remain on disk and are reachable via Read and Bash. Skills are model-invoked; pair them with a tight allowedTools list (and dontAsk where appropriate) so an invoked skill cannot use more tools than intended. The allowed-tools frontmatter in SKILL.md does not apply through the SDK; control tool access with the main allowedTools option.The store is a mirror, not a replacement: the CLI writes locally first, then append() forwards. sessionStore cannot combine with persistSession: false or enableFileCheckpointing. Mirror writes are best-effort; a failed append emits a system mirror_error message and the query continues, so monitor for mirror_error to detect store data loss. Deleting the main session key must cascade to subagent subkeys; implement delete carefully or treat the backend as append-only.Production agents execute real tools; start with read-only allowlists and expand after review. Do not mount host credentials into SDK processes—use a proxy pattern for secrets. Unhandled tool exceptions can hang or crash query loops; return structured errors from handlers.
Privacy notesBundled MCP servers and hooks can access local files and external services; confirm what a plugin's components reach before deploying it. Plugin skills and commands are sent to the model as available capabilities; keep secrets out of plugin definitions. Loading a plugin from a path you do not control is a supply-chain decision; prefer plugins you have reviewed and pinned.Skill descriptions are loaded so the model can decide when to use them; keep sensitive workflow detail and secrets out of descriptions. Skills sourced from outside your project run their instructions in your sessions; review them before enabling. Skill content is sent to the model provider when a skill is invoked; treat it like any other prompt content.Transcripts contain the full conversation, tool inputs, and outputs; store them in a backend you govern with your own encryption, access controls, and retention. The SDK never deletes from your store; retention (TTLs, S3 lifecycle, scheduled cleanup) is your responsibility. Keep backend credentials out of code; pass a pre-configured client to the adapter so you control TLS, region, and pooling.Agent transcripts and tool I/O are sent to the model provider unless using a private deployment. Log pipelines may retain prompts—scrub PII before shipping logs to third-party observability. Multi-tenant hosts must isolate session storage per tenant.
Prerequisites
  • The Claude Agent SDK installed for Python or TypeScript.
  • One or more plugin directories on the local filesystem (the parent of skills/, agents/, hooks/, or .claude-plugin/).
  • For marketplace plugins, the plugin downloaded locally so you can provide a path.
  • The Claude Agent SDK installed for Python or TypeScript.
  • SKILL.md files in .claude/skills/ (project) or ~/.claude/skills/ (user).
  • A cwd that points at or below the directory containing .claude/skills/, within the same repository.
  • The Claude Agent SDK installed for Python or TypeScript.
  • An external backend (S3, Redis, a database) and its client library.
  • A working-directory layout where local transcript writes are available (the store mirrors them).
  • Node.js 18+ or Python 3.10+ for TypeScript or Python Agent SDK packages.
  • Anthropic API credentials or approved provider configuration for CI and prod.
  • A bounded first use case with explicit tool allowlists—not an open-ended chatbot.
  • Staging environment for soak testing before customer traffic.
Install
Config
Citations
ClaimUnclaimedUnclaimedUnclaimedUnclaimed
Open 4 picks in the interactive comparison tool

Related guides

Signals

Loading live community signals…

More like this, weekly

A short, calm digest of reviewed Claude resources. Unsubscribe any time.