← Back to Skills Marketplace
ajspig

Long-Term Memory (Honcho)

by ajspig · GitHub ↗ · v1.0.4 · MIT-0
cross-platform ⚠ suspicious
1333
Downloads
0
Stars
2
Active Installs
5
Versions
Install in OpenClaw
/install honcho-setup
Description
Install the @honcho-ai/openclaw-honcho plugin and run initial setup. Runs `openclaw plugins install`, `openclaw honcho setup` (which prompts for your API key...
README (SKILL.md)

Honcho Setup

Install the @honcho-ai/openclaw-honcho plugin. The plugin includes openclaw honcho setup, which handles API key configuration and migration of any legacy memory files to Honcho.

⚠️ DATA UPLOAD WARNING: openclaw honcho setup (run in Step 2) will offer to upload your workspace memory files (USER.md, MEMORY.md, IDENTITY.md, memory/, canvas/, SOUL.md, AGENTS.md, BOOTSTRAP.md, TOOLS.md) to an external API. HEARTBEAT.md is excluded. By default, data is sent to api.honcho.dev. For self-hosted instances, data is sent to your configured HONCHO_BASE_URL. The setup command will show exactly which files will be uploaded and ask for explicit confirmation before proceeding.

No local files are deleted, moved, or modified. Originals remain exactly in place.

Step 1: Install the Plugin

Install the Honcho plugin using the OpenClaw plugin system. Use this exact command — do not install @honcho-ai/sdk directly or use npm install in the workspace.

openclaw plugins install @honcho-ai/openclaw-honcho

Then enable it:

openclaw plugins enable openclaw-honcho

Verify the plugin loaded without errors. If the gateway logs show Cannot find module '@honcho-ai/sdk', install dependencies manually:

cd ~/.openclaw/extensions/openclaw-honcho && npm install

This is a known issue with the OpenClaw plugin installer not running dependency resolution for plugin packages.

Step 2: Run Setup

Run the setup command that ships with the plugin:

openclaw honcho setup

This command will:

  1. Prompt interactively for your Honcho API key
  2. Write configuration to ~/.openclaw/openclaw.json
  3. Scan for legacy memory files and offer to migrate them to Honcho

Follow the prompts. Migration is optional — if you have no legacy files or want to skip, you can skip the upload step.

For managed Honcho, you need an API key from https://app.honcho.dev. For self-hosted instances, set HONCHO_BASE_URL to your instance URL and the API key is optional.

Step 3: Restart the Gateway

openclaw gateway restart

Step 4: Confirm

Verify the plugin is active by checking gateway logs or running:

openclaw honcho status

Honcho memory is now active.

Ongoing behavior after setup: Once enabled, the plugin will persistently observe conversations in this workspace and send conversation data to api.honcho.dev (or your configured HONCHO_BASE_URL) to build and retrieve memory. This is ongoing network activity that continues across sessions. Memory is made available via honcho_recall, honcho_search, honcho_profile, and related tools. To stop this behavior, disable the plugin with openclaw plugins disable openclaw-honcho.

Security & Privacy Disclosure

This skill has been designed with transparency and safety as priorities. Below is a complete disclosure of what this skill does:

What This Skill Does

This skill runs three commands: openclaw plugins install, openclaw honcho setup, and openclaw gateway restart. The data upload and file access described below is performed by openclaw honcho setup, not by this skill directly.

Data Upload

  • uploaded_content: USER.md, MEMORY.md, IDENTITY.md, all files under memory/, all files under canvas/, SOUL.md, AGENTS.md, BOOTSTRAP.md, TOOLS.md — uploaded by openclaw honcho setup during migration
  • not_uploaded: HEARTBEAT.md — excluded by design, never read or uploaded
  • Where it goes: By default to api.honcho.dev (managed Honcho cloud service). For self-hosted instances, to your configured HONCHO_BASE_URL
  • User control: openclaw honcho setup always requires explicit interactive confirmation before any upload, even when HONCHO_API_KEY is pre-set in the environment. You will see the exact list of files and the destination URL
  • Purpose: Migrating file-based memory system to Honcho API for AI agent personalization and memory

File Modifications

  • Config written: openclaw honcho setup writes API key and config to ~/.openclaw/openclaw.json
  • Workspace files: Never modified — originals remain exactly in place
  • HEARTBEAT.md: Never read or uploaded — excluded by design

Credentials

  • HONCHO_API_KEY: Required only for managed Honcho (api.honcho.dev). Not required for self-hosted instances. Stored in ~/.openclaw/openclaw.json by the setup command.
  • No other credentials: This skill does not access, read, or transmit any other credentials or secrets

Network Access

  • Managed mode: Connects to api.honcho.dev (Honcho cloud service)
  • Self-hosted mode: Connects to your configured HONCHO_BASE_URL (e.g., http://localhost:8000)

Ongoing Behavior After Setup

  • Persistent observation: Once enabled, the plugin observes all conversations in the workspace and transmits conversation data to the configured Honcho endpoint on an ongoing basis
  • Network activity: Continues across sessions as long as the plugin is enabled — this is not a one-time migration
  • How to stop: Run openclaw plugins disable openclaw-honcho to stop all observation and network activity

Open Source

  • Honcho SDK: Open source at https://github.com/plastic-labs/honcho
  • Plugin code: Available at ~/.openclaw/extensions/openclaw-honcho after installation
  • This skill: You are reading the complete skill instructions - there is no hidden behavior
Usage Guidance
This skill does what it says: it installs an Honcho plugin and offers to upload many workspace/memory files to a remote Honcho API (api.honcho.dev by default). Before installing, confirm you trust Honcho (managed service) or set HONCHO_BASE_URL to a self-hosted instance you control. Pay attention to the interactive prompts — the setup requires explicit confirmation before any upload. Be aware the plugin will write your API key to ~/.openclaw/openclaw.json and will continue to observe and transmit conversation data across sessions until you disable it (openclaw plugins disable openclaw-honcho). Also note minor metadata inconsistencies in the skill package (required binaries listed in SKILL.md but not in the registry summary, and a version mismatch in _meta.json); inspect the installed plugin code under ~/.openclaw/extensions/openclaw-honcho if you want to verify behavior before enabling.
Capability Analysis
Type: OpenClaw Skill Name: honcho-setup Version: 1.0.4 The skill automates the installation of the Honcho AI plugin, which performs a bulk upload of sensitive workspace files (including USER.md, IDENTITY.md, AGENTS.md, and memory directories) to an external service (api.honcho.dev). While the documentation in SKILL.md is highly transparent about this behavior and the requirement for user confirmation, the inherent capability for large-scale data exfiltration and persistent conversation monitoring across sessions constitutes a significant security and privacy risk.
Capability Assessment
✓ Purpose & Capability
The name/description (Long-Term Memory / Honcho) align with the instructions: installing an OpenClaw Honcho plugin, running its setup, and migrating/uploading workspace memory files to the Honcho API. Optional env vars (HONCHO_API_KEY, HONCHO_BASE_URL) and writing config to ~/.openclaw/openclaw.json are consistent with integrating a remote memory service. Minor metadata mismatch: registry 'Requirements' reported none, while SKILL.md metadata lists node and npm as required binaries.
ℹ Instruction Scope
SKILL.md explicitly instructs three commands (plugin install, plugin setup, gateway restart) and clearly documents which files will be uploaded (USER.md, MEMORY.md, memory/, canvas/, SOUL.md, IDENTITY.md, AGENTS.md, BOOTSTRAP.md, TOOLS.md) and that HEARTBEAT.md is excluded. The upload and persistent observation behavior is called out and requires explicit interactive user confirmation before upload. This scope is broader (it transmits many personal/workspace files) but is consistent with the stated purpose.
✓ Install Mechanism
This is an instruction-only skill with no install spec or bundled code — the skill tells the user to run the OpenClaw plugin installer, which is the platform's normal mechanism. No direct downloads or archive extraction are performed by the skill itself. The fallback 'npm install' guidance (in case the plugin installer does not install dependencies) is reasonable. No high-risk install URLs or extract operations are present in the skill text.
ℹ Credentials
The skill does not demand unrelated credentials; it optionally uses HONCHO_API_KEY for managed Honcho and HONCHO_BASE_URL for self-hosting. It will prompt for an API key and write it to ~/.openclaw/openclaw.json — this is proportional to connecting to a remote memory service but is sensitive. Users should understand that providing an API key or using managed Honcho enables uploads of workspace files to an external endpoint.
ℹ Persistence & Privilege
always:false (default) and autonomous invocation is allowed (platform default). The plugin, once enabled, will persistently observe conversations and send data across sessions — this persistent network behavior is consistent with a memory plugin but increases privacy risk. The skill does not request system-wide privileges or attempt to modify other skills' configs beyond writing its own plugin config.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install honcho-setup
  3. After installation, invoke the skill by name or use /honcho-setup
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.4
honcho-setup 1.0.4 - Clarified that data upload via `openclaw honcho setup` always requires explicit interactive user confirmation, even when `HONCHO_API_KEY` is set. - Updated documentation to warn that, after setup, the plugin persistently observes conversations and transmits data to Honcho on an ongoing basis. - Expanded instructions on disabling the plugin to stop ongoing data transmission. - No changes to files or core functionality; documentation only.
v1.0.3
honcho-setup 1.0.3 changelog: - Added _meta.json file to the repository. - Updated documentation to clarify: workspace and memory files are never deleted, moved, or modified—only uploaded with user confirmation. - HEARTBEAT.md is now excluded from upload/migration. - Setup process streamlined to rely on openclaw honcho setup (API key prompt, config written to ~/.openclaw/openclaw.json). - Instructions and metadata simplified for improved clarity regarding security, privacy, and migration steps.
v1.0.2
honcho-setup v1.0.2 - Clarified and strengthened data handling disclosures in documentation (SKILL.md). - Explicitly details that workspace files are uploaded to external destinations (api.honcho.dev or self-hosted) and requires user confirmation. - Documents that only HONCHO_API_KEY is read from ~/.openclaw/.env, and no other sensitive values are accessed. - Adds a data_handling section specifying uploaded file types, external endpoints, and confirmation requirements. - No functional or code changes; documentation improvements only.
v1.0.1
honcho-setup 1.0.1 - Added SKILL.md metadata fields specifying environment variables, required binaries, archive location, and network access. - Clarified that uploaded workspace content may be sent to api.honcho.dev or the user’s self-hosted endpoint. - Expanded environment variable documentation, including HONCHO_API_KEY (optional if self-hosted) and HONCHO_BASE_URL. - Noted binary requirements (node, npm; optionally git, docker, docker-compose). - Updated skill description and documentation for improved clarity and transparency regarding migration, archiving, and network activity. - No changes to code or behavior—documentation and metadata only.
v1.0.0
honcho-setup 1.0.0 - Initial release of the honcho-setup skill. - Installs the @honcho-ai/openclaw-honcho plugin and migrates legacy workspace memory files (USER.md, MEMORY.md, IDENTITY.md, memory/, canvas/) to Honcho. - Supports both managed (cloud API key) and self-hosted Honcho instances; verifies connection before proceeding. - Detects legacy files and asks for user confirmation before uploading, archiving, or removing any files. - Archives originals and updates workspace documents (SOUL.md, AGENTS.md, BOOTSTRAP.md) to reference Honcho tools after successful upload. - Prevents any file modification or removal if connection or upload fails.
Metadata
Slug honcho-setup
Version 1.0.4
License MIT-0
All-time Installs 2
Active Installs 2
Total Versions 5
Frequently Asked Questions

What is Long-Term Memory (Honcho)?

Install the @honcho-ai/openclaw-honcho plugin and run initial setup. Runs `openclaw plugins install`, `openclaw honcho setup` (which prompts for your API key... It is an AI Agent Skill for Claude Code / OpenClaw, with 1333 downloads so far.

How do I install Long-Term Memory (Honcho)?

Run "/install honcho-setup" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Long-Term Memory (Honcho) free?

Yes, Long-Term Memory (Honcho) is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Long-Term Memory (Honcho) support?

Long-Term Memory (Honcho) is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Long-Term Memory (Honcho)?

It is built and maintained by ajspig (@ajspig); the current version is v1.0.4.

More Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 Comments