personal-identity

Personal Identity & Brand

This skill governs the identity and brand modules of the Personal Brain OS. It covers the files that define how you write, how you're perceived, and how you choose to show up professionally. These files are foundational — the content pipeline, engineering, and operations modules all pull from identity as their source of truth.

The identity module reads only static narrative markdown and YAML config files from ~/brain/identity/ and ~/brain/brand/. There are no entity directories to glob, no observation masking needed, and no JSONL data here. All reads are deliberate and file-specific.

---

When to Use This Skill

Load this skill for any of the following:

• Content creation — generating any post, article, thread, newsletter, or email the user will publish or send
• Voice check — reviewing whether a draft matches the user's established voice
• Bio request — producing a bio for any platform, length, or context
• Brand positioning — clarifying audience, positioning, or messaging strategy
• Anti-patterns review — checking whether a draft uses any banned phrases or structural clichés
• Values clarification — surfacing ranked values when the user is making a decision or trade-off
• Identity bootstrapping — helping the user fill in voice.md, anti-patterns.md, bio-variants.md, values.yaml, or brand/positioning.md for the first time

Do NOT load this skill for engineering decisions, contact lookups, goal tracking, or journaling — those belong to other modules.

---

Files in This Module

All paths relative to ~/brain/:

File	Purpose	When to Load
identity/voice.md	Primary voice reference — tone, vocabulary, rhythm, platform adaptations	Required before any content output
identity/anti-patterns.md	Writing patterns, phrases, and structures to avoid	Before presenting any draft
identity/values.yaml	Ranked personal values with non-negotiables and operating principles	When the user faces a decision, trade-off, or values question
identity/bio-variants.md	Short, medium, and full bio variants plus platform overrides	When the user requests a bio
brand/positioning.md	Brand positioning statement, core audience, pillars, unique angle	When the user asks about brand, positioning, or audience

Setup check: If any of these files is missing, inform the user and suggest using the /bootstrap command to create it from the reference template in skills/personal-identity/references/.

---

Behavioral Rules

These rules operationalize AGENT.md Rules #1 and #2 for this module. They are non-negotiable for all content tasks.

Rule PI-1 — Voice First (always)

Before generating any content output, read ~/brain/identity/voice.md in full.

• Content includes: posts, threads, newsletters, blog drafts, emails, meeting summaries, idea expansions, bios, and any other text the user will read, publish, or send
• If voice.md is absent: stop, inform the user, suggest running /bootstrap personal-identity
• Do not proceed with content generation if voice.md cannot be read

Rule PI-2 — Anti-Patterns Check (before every draft)

Before presenting any draft, read ~/brain/identity/anti-patterns.md and run a mental diff against the draft.

• If the draft uses any banned phrase, structural pattern, or tonal register listed in anti-patterns.md: rewrite before presenting
• If anti-patterns.md is absent: skip this check and proceed (no anti-patterns defined yet is a valid state)
• Never present a first draft that violates a known anti-pattern

Rule PI-3 — Bio Selection Logic

When the user asks for a bio:

1. Check the request for length signal: "tweet", "short", "one-liner", "140 chars", "280 chars" → Short Bio; "headline", "medium", "LinkedIn" → Medium Bio; "speaker", "full", "conference", "about page", "3–5 sentences" → Full Bio
2. If no length signal: default to Medium Bio, then offer Short and Full as alternatives
3. Load ~/brain/identity/bio-variants.md → return the matching variant verbatim
4. If a platform is specified (e.g., "for GitHub"), check the Platform-Specific Overrides table in bio-variants.md first

Rule PI-4 — Brand Queries

When the user asks about brand, positioning, audience, or messaging strategy:

1. Load ~/brain/brand/positioning.md
2. If the question is about values underlying brand decisions, also load ~/brain/identity/values.yaml
3. Never conflate voice (how you say things) with positioning (why you're the right choice for your audience) — keep them distinct in your response

Rule PI-5 — File Loading Order

For content tasks, always load in this order:

1. identity/voice.md          (required — sets the voice anchor)
2. identity/anti-patterns.md  (optional — check if present)
3. task-specific file         (bio-variants.md, brand/positioning.md, values.yaml as needed)

Never load task-specific files before voice.md. Never reverse this order.

Rule PI-6 — Values as a Tiebreaker

When the user is deciding between options and asks which aligns with their values:

1. Load ~/brain/identity/values.yaml
2. Identify which option better satisfies the highest-ranked non-negotiable values
3. Flag any option that crosses a red_lines boundary
4. Present the analysis, then let the user decide

---

Reference Templates

These files live in skills/personal-identity/references/ and serve as the starting templates for bootstrapping the user's ~/brain/identity/ and ~/brain/brand/ directories. Use /bootstrap to copy them into the vault.

Template	Copies To	Purpose
voice-template.md	~/brain/identity/voice.md	Tone, vocabulary, rhythm, platform adaptations, examples
anti-patterns-template.md	~/brain/identity/anti-patterns.md	Banned phrases, structural anti-patterns, tonal limits
bio-variants-template.md	~/brain/identity/bio-variants.md	Short / medium / full bio variants + platform overrides
brand-template.md	~/brain/brand/positioning.md	Positioning statement, audience, pillars, unique angle
values-template.yaml	~/brain/identity/values.yaml	Ranked values, operating principles, decision heuristics

---

Version Notes

• v1.0.0 — Phase 2A: initial personal-identity sub-skill. Covers identity + brand modules. No entity directories — static file reads only. Behavioral rules PI-1 through PI-6 implement AGENT.md Rules #1 and #2 at the module level.