Bolta.draft.post

Goal

Create one post in the user's Brand Voice (or Founder Voice), optionally using a template, and route it safely through the V2 state machine.

Voice Profile Modes

Brand Voice

• Company/brand speaking
• Focus on tone, style rules, vocabulary
• Professional and consistent across content

Founder Voice (Advanced)

• Founder speaking personally
• Includes founder_signals: beliefs, lessons, contrarian opinions, scars, backstory moments
• More authentic, personal, thought-leadership focused
• founder_signals are injected as constraints in content generation

Which Agent Types Use This

• content_creator — Primary use case (generates posts as part of jobs)
• custom — User-defined agents with content creation enabled

Hard Rules

1. Always call bolta.get_workspace_policy + bolta.get_my_capabilities first.
2. Safe Mode enforcement (workspace-level):
   • If Safe Mode ON: ALL content goes to Inbox for approval, regardless of role
   • If Safe Mode OFF: Creator/Viewer still route to Inbox; Editor/Admin can schedule directly
3. If actor role does not include the required capability for an action (schedule/publish), MUST NOT attempt it.
4. If requested_action is omitted: default to send_to_inbox (if Safe Mode ON or role is Creator/Viewer), otherwise draft_only.
5. Never publish directly without explicit approval.
6. Never schedule unless capability includes posts:schedule.

How Voice Profiles Constrain Generation

Tone Constraints (all modes)

Content generation respects tone scores (playful, professional, direct, thoughtful):

• Tone scores guide sentence structure, vocabulary choice, emoji usage
• Direct tone (9/10) = fewer hedges, more imperatives
• Thoughtful tone (7/10) = more context, nuance, depth

Style Rules (all modes)

• dos: Patterns to follow (lead with pain points, use short sentences)
• donts: Patterns to avoid (no corporate jargon, avoid hedging)
• styleKeywords: Vocabulary bias (emphasize direct, confident, specific)

Founder Signals (founder mode only)

When speakerMode: "founder" and founderSignals provided:

• Beliefs: Constraints that shape main arguments
  • Example: If belief is "Ship fast beats planning," posts favor speed over perfection
• Lessons: Constraints that shape advice given
  • Example: If lesson is "Users teach more than docs," content leads with user validation
• Contrarian Opinions: Constraints that enable pushback
  • Example: If contrarian is "Engagement metrics are vanity," posts optimize for substance over reach
• Scars: Constraints that enable vulnerability
  • Example: If scar is "Built feature no one asked for," posts emphasize user research
• Backstory Moments: Constraints that enable narrative coherence
  • Example: If backstory is "Spent years ghostwriting," posts demonstrate writing from real experience

Founder Signal Injection Example:

Prompt: "Write about validating product ideas"
Voice Profile:
  - speakerMode: "founder"
  - beliefs: ["Ship fast", "Users teach more than docs"]
  - lesson: "Build what users ask for"
  - scar: "Built feature no one wanted"

Generated Content Characteristics:
- Opens with user validation (belief: users teach)
- Mentions framework for speed (belief: ship fast)
- Includes personal example of failure (scar: built wrong feature)
- Gives specific, actionable advice (lesson learned)
- Tone: Direct + thoughtful (8/10 and 7/10)

Generated: "Most founders waste months on features no one asked for—here's my 3-question filter..."
(This naturally demonstrates the beliefs/lessons/scars without being forced)

Steps

1. Check policy & capabilities

• bolta.get_workspace_policy(workspace_id) → extract safe_mode (boolean)
• bolta.get_my_capabilities(workspace_id) → check role capabilities

2. Load context for consistency

• bolta.get_voice_profile(voice_profile_id) → tone, style rules, banned phrases
  • If founder mode: Extract founder_signals for generation constraints
• bolta.list_recent_posts(account_ids, limit=5) → avoid repetition, maintain consistency
• If agent_id provided: bolta.recall(agent_id, "content_preferences") → retrieve agent memory

3. Generate content

• If template_id provided:
  • bolta.list_templates(workspace_id) (if needed for validation)
  • bolta.render_template(template_id, {prompt, voice_profile_id, context})
  • Use rendered text as the post body
• Else:
  • Generate body using voice profile constraints:
    • Tone scores (all modes)
    • Style rules (all modes)
    • Founder signals (if founder mode)
    • Custom rules and keywords
  • Check recent posts to avoid repetition

4. Create the draft

• bolta.draft_post({workspace_id, voice_profile_id, account_ids, content: body, status: "draft"})
• Returns post_id

5. Route based on governance

• If Safe Mode ON OR role is Creator/Viewer:

  • Route to Inbox: post transitions to inbox state
  • inbox_item_id created automatically by system
  • final_state = inbox

• If Safe Mode OFF AND role is Editor/Admin:

  • requested_action == draft_only → leave in draft
  • requested_action == send_to_inbox → route to inbox
  • requested_action == schedule → bolta.schedule_post(post_id, suggested_time) → final_state = scheduled
  • requested_action == publish → bolta.publish_post(post_id) → final_state = published

6. Store learnings (if agent context)

• If successful and agent_id provided:
  • bolta.remember(agent_id, "last_successful_hook_style", hook_type_used)
  • Update agent memory with what worked

Output

Return:

{
  "post_id": "uuid",
  "final_state": "inbox",
  "inbox_item_id": "uuid",
  "notes_for_reviewer": "AI draft ready for review",
  "warnings": []
}

Failure Handling

• If any tool call returns unauthorized/forbidden: downgrade to inbox (if possible) or leave in draft.
• Never retry publish/schedule more than once.
• On failure, create inbox item with error context for human review.
• Log to Run record if executed as part of a Job.

V2 Agent Context

When agents use this skill:

Content Creator reasoning:

"I need to draft a post about our new feature. First, let me check the voice profile to match tone. Then I'll review recent posts to avoid repetition. I see we posted about pricing yesterday, so I'll focus on product benefits instead."

Key differences from V1:

• V1: Template fill → post
• V2: Agent reasons about context → chooses tools → drafts content → learns from feedback

Example execution trace:

Turn 1: get_voice_profile() → "Direct, specific, no jargon"
Turn 2: list_recent_posts() → 2 about pricing (avoid)
Turn 3: recall() → "Educational > promotional 3:1"
Turn 4: draft_post() → Created with educational angle
Turn 5: remember() → Store "educational_format_used"

Example Usage

Scenario 1: Job Execution (Safe Mode ON)

{
  "workspace_id": "uuid",
  "agent_id": "uuid",
  "job_id": "uuid",
  "voice_profile_id": "uuid",
  "account_ids": ["twitter-uuid"],
  "prompt": "Write a post about our new feature launch",
  "requested_action": "send_to_inbox"
}

Result: Post created → routed to Inbox → awaits approval → then scheduled/published

Scenario 2: Manual Draft (Safe Mode OFF, Editor role)

{
  "workspace_id": "uuid",
  "voice_profile_id": "uuid",
  "account_ids": ["linkedin-uuid"],
  "prompt": "Share our hiring announcement",
  "requested_action": "schedule",
  "suggested_time": "2026-02-21T09:00:00Z"
}

Result: Post created → immediately scheduled for 9am tomorrow (bypasses inbox)