← Back to Skills Marketplace
175
Downloads
0
Stars
0
Active Installs
8
Versions
Install in OpenClaw
/install remember-when-skill
Description
An intelligent digital archivist skill for OpenClaw agents. It allows agents to monitor chat groups, capture memorable content (text, photos, videos, interes...
Usage Guidance
This skill is internally consistent, but before installing: (1) verify and install the remember-when CLI from a trusted source (the skill requires that binary but does not provide it), (2) review and approve any agent config changes (AGENTS.md / TOOLS.md / HEARTBEAT.md) the skill asks you to make — these grant the agent broad read access to chat streams and write access to ~/.remember-when, (3) confirm the storage location and retention policy for archived data and who can access those files, (4) test the skill in a sandboxed agent or isolated account first to confirm it truly performs only local operations (SKILL.md claims no external network calls), and (5) if you need stricter privacy controls, limit the agent's chat-read scope or require explicit user confirmation before each archival action. If you want a tighter assessment, provide the remember-when CLI source/binary or the exact agent integration points so I can check for any hidden external endpoints or required permissions that are not documented.
Capability Analysis
Type: OpenClaw Skill
Name: remember-when-skill
Version: 1.2.3
The 'remember-when-skill' is a digital archivist tool designed to monitor chat groups and archive media/text to local storage via a dedicated CLI. While it requests high-level autonomy—including instructions for the agent to modify its own internal configurations (AGENTS.md, TOOLS.md) for persistence and heartbeat routines—these actions are transparently documented and aligned with the stated purpose of proactive archiving. Crucially, the SKILL.md includes a 'Validation Questions API' that mandates user confirmation before bulk operations or major changes, and there is no evidence of data exfiltration, obfuscation, or unauthorized access to sensitive system files like SSH keys or environment variables.
Capability Tags
Capability Assessment
Purpose & Capability
Name/description require the remember-when CLI and local persistence; SKILL.md only instructs using that CLI and local files (~/.remember-when). There are no unrelated cloud credentials, binaries, or installs requested, so the declared capabilities match the requested access.
Instruction Scope
Instructions direct the agent to 'listen to all messages', be proactive, read group histories, and modify internal agent config (AGENTS.md, TOOLS.md). That is within scope for an archivist but has privacy/coverage implications: it requires broad read access to chat streams and the ability to alter agent config. The instructions explicitly ask the agent to perform autonomous archival, daily audits, and re-reading history — all consistent with purpose but warrant user review of permissions and audit behaviour.
Install Mechanism
This is an instruction-only skill with no install spec; it only requires an existing remember-when binary on PATH. README documents installation of the skill itself separately. No downloads or archive extraction are requested by the skill package.
Credentials
No environment variables, credentials, or external API keys are requested. The only required access is filesystem persistence (~/.remember-when/) and access to message streams (implied). These are proportionate to a local archivist that saves chat content.
Persistence & Privilege
The skill does not request always: true and uses normal autonomous invocation. It instructs the operator to grant the CLI and the agent permissions to write to a home-directory storage and to alter agent internal config files; this is normal for an autonomous archivist but you should review and approve those config changes manually.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install remember-when-skill - After installation, invoke the skill by name or use
/remember-when-skill - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.2.3
- Added support for cross collections, enabling grouping of entries from different days within the same group under a thematic umbrella.
- Updated SKILL.md documentation with instructions for creating, managing, and using cross collections and their CLI commands.
- Clarified agent configuration requirements for storage root access.
- Minor documentation updates for improved clarity.
v1.2.2
## remember-when-skill 1.2.2 Changelog
- Updated `package.json` version to 1.2.2.
- No code or documentation changes; this release updates version metadata only.
v1.2.1
- Bumped version in package.json from 1.2.0 to 1.2.1.
- No functional changes; documentation and skill logic remain unchanged.
v1.2.0
- Updated package metadata in package.json.
- No functional or documentation changes to the skill itself.
v1.1.8
- Minor documentation updates and clarifications
- Improved language regarding file access and update operations in configuration instructions
- Changed certain terms for clarity (e.g., "read/write access" to "persistence access"; "overwrite" to "update")
- No changes to features or functionality; this is a documentation and language refinement release
v1.1.7
**Major update: Introduces proactive archiving, expanded entry types, and detailed agent guidance.**
- Added proactive archiving workflow: the agent now detects and offers to archive valuable content without waiting for user instruction.
- Expanded supported entry types to include `interest_point` (places, locations) and clarified valid types.
- Introduced step-by-step workflows for context buffering, interview (pre-flight check), duplicate prevention, and audit loop.
- Updated agent configuration requirements, emphasizing context buffers, permissions, and heartbeat routines for multimedia intake.
- Enhanced validation steps and confirmation prompts for safer and more autonomous operation.
- Version bump to 1.2.0.
v1.1.6
Version 1.1.6
- Added a new Security & Privacy Statement outlining local-only data processing, storage location, and absence of credential requirements.
- Simplified archivist instructions and workflow steps for greater clarity.
- Reduced instruction verbosity; removed detailed prompt guidance and examples.
- Minor updates to documentation formatting and requirements.
v1.1.5
- Enhanced skill description and documentation to clarify group monitoring, content capture, and summarization features.
- Updated archivist instructions, detailing what to capture, workflows, commands, and privacy practices.
- Clearly documented required tools and dependencies for easy setup.
- Added guidance for maintaining and auditing group context and recorded memories.
Metadata
Frequently Asked Questions
What is Remember When Skill?
An intelligent digital archivist skill for OpenClaw agents. It allows agents to monitor chat groups, capture memorable content (text, photos, videos, interes... It is an AI Agent Skill for Claude Code / OpenClaw, with 175 downloads so far.
How do I install Remember When Skill?
Run "/install remember-when-skill" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Remember When Skill free?
Yes, Remember When Skill is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Remember When Skill support?
Remember When Skill is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Remember When Skill?
It is built and maintained by Eric Mora (@ericmora); the current version is v1.2.3.
More Skills