Digital Brain Personal OS
Manage content, relationships, knowledge, and goals through structured files and AI.
The documentation specifies append-only storage, history preservation, and module boundaries, which reduce accidental deletion risk. However, it handles contacts, interactions, goals, and personal-brand data without defining least privilege, sensitive-data handling, user confirmation, data flows, external side effects, or recovery/rollback, so points are deducted.
The routing guidance and several Python scripts show plausible happy paths, with some handling for missing files and invalid JSON. However, referenced key scripts such as weekly_review.py and metrics_snapshot.py are not supplied, malformed JSON is silently skipped, and input ambiguity, write failures, and script exceptions lack diagnosable feedback. Static evidence is insufficient and the score remains at or below the required 10-point ceiling.
Trigger phrases, target users, and primary scenarios are reasonably clear across content, contacts, meetings, and reviews. Boundaries for non-fit cases, input constraints, platform paths, and alternate data layouts are not sufficiently stated. Chinese-language support is not documented, but the core function does not entirely depend on overseas online services, so no additional mainland-network deduction is applied.
The skill provides progressive disclosure, module structure, format conventions, examples, version, author, license, and update metadata. Nevertheless, referenced modules cannot be fully verified from the supplied material, script inventories are inconsistent, creation and update dates conflict, and maintenance ownership, changelog, installation prerequisites, and systematic troubleshooting are incomplete, so points are deducted.
Content generation, contact lookup, and meeting-preparation workflows describe directly usable outputs, and the append-only file model has practical value. However, results depend heavily on users populating identity and data files, the supplied implementation is incomplete, and quality and quantified benefits are primarily asserted. Static review cannot confirm end-to-end completion of the core tasks, so the score is conservative.
The material includes code, workflows, and repository CI, but the shown CI mainly validates the researcher system and other examples rather than Digital Brain’s key paths. Quantified token and outcome claims lack skill-specific tests or third-party execution evidence, and external references were not independently verified in this review, so only limited points are awarded.
- Define access boundaries for contact, interaction, and personal-brand data, and require confirmation before writes or archival changes.
- Provide or reconcile weekly_review.py, metrics_snapshot.py, and the referenced module files to prevent documentation/package mismatches.
- Do not treat silently skipped invalid JSON as success; report the file, line, and cause, and provide a recovery path for write failures.
- The quantified benefits and production-readiness claims are not supported by skill-specific tests.
What does this skill do, and when should you use it?
Digital Brain is a personal operating-system skill for founders, content creators, and tech-savvy professionals. It covers content creation, personal brand management, contact and relationship workflows, meeting preparation, knowledge organization, weekly reviews, and goal tracking. Its three-level progressive disclosure model loads module instructions and data only when needed. The system uses local Markdown, JSONL, YAML, and XML files and includes automation scripts for reviews, content ideas, stale contacts, and idea-to-draft workflows.
The skill reads local files in the identity, content, knowledge, network, and operations modules and produces voice-aligned drafts, relationship and meeting briefs, review outputs, and goal-update guidance. Its content pipeline uses ideas.jsonl, drafts, posts.jsonl, and calendar.md; relationship workflows use contacts.jsonl and interactions.jsonl. Weekly reviews can run python agents/scripts/weekly_review.py and check stale_contacts.py. JSONL records are append-only, with archived status used instead of deletion.
- A founder wants to draft a public-building post that matches their established voice and brand positioning.
- A professional is preparing for a meeting and needs a brief built from contact history, interactions, and pending tasks.
- A content creator wants to capture raw ideas in ideas.jsonl and develop them into drafts and publication records.
- An individual wants a weekly review covering metrics, goal progress, stale contacts, and the next content plan.
- A researcher wants to organize bookmarks, research, and learning materials for later use in content work.
What are this skill's strengths and limitations?
- Connects content, brand, knowledge, relationships, meetings, tasks, and goals in one operating model.
- Progressive disclosure and module isolation help limit unnecessary context loading.
- Append-only JSONL records preserve history for performance and relationship retrospectives.
- Provides concrete workflows for content creation, meeting preparation, and weekly reviews.
- Users must maintain the local folder structure, identity materials, contact records, and historical data.
- The source does not document migration, access control, backup, or multi-user collaboration.
- Review and contact-check workflows require Python and the expected local files; the scripts' implementation details are not included in SKILL.md.
- The source does not provide a dedicated test suite or cross-platform validation results for this example.
How do you install this skill?
Clone the repository, then copy examples/digital-brain-skill/ into a host skill directory while preserving the full folder structure. For example: mkdir -p .claude/skills; cp -R examples/digital-brain-skill .claude/skills/digital-brain. The folder can also be copied into .codex/skills/. The source does not document any additional package-installation step.
How do you use this skill?
Place the folder in an Agent Skills-compatible host and use prompts such as “write a post about AI agents,” “prepare for meeting with Sarah Chen,” or “weekly review.” Before creating content, read identity/voice.md, then check brand.md and previous posts. For meeting preparation, read contacts.jsonl, interactions.jsonl, and operations/todos.md. Load module files on demand rather than loading the entire data set at once.