prospecting-signal-context-builder

name

prospecting-signal-context-builder

description

Build product-agnostic GTM signal context files for signal detection. Use when Codex needs to turn user-provided text or files into a normalized `signal-context.md` containing product context, sector context, ICP, competitors, alternatives, pain points, trigger events, disqualification rules, signal criteria, search seeds, and engagement constraints for use by the prospecting-signal-detector skill.

Signal Context Builder

Use this skill to create or update signal-context.md for prospecting-signal-detector.

This skill is product, project, and sector agnostic. All product context, sector context, ICP, competitors, alternatives, signal criteria, and engagement constraints must come from user-provided text or files.

Inputs

Accept source material from:

• The user's current message or pasted notes.
• Workspace files named by the user.
• Product briefs, positioning docs, website copy, sales notes, ICP notes, CRM exports, competitor notes, market research, call transcripts, campaign plans, or existing signal-context.md.
• Links or live research only when the user asks for current external evidence.

If source material is missing or too thin, ask for the smallest missing high-impact input. Do not invent product claims, sectors, ICPs, competitors, alternatives, or signal definitions.

Output

Write a human-editable markdown file named signal-context.md at the workspace root unless the user asks for another path.

Use references/context-template.md as the output structure. Keep unknowns explicit with Unknown or Needs input, and include a short Open Questions section when important gaps remain.

Workflow

1. Gather source material.
   • Read user-named files first.
   • Preserve the source-of-truth order the user provides.
   • If existing signal-context.md exists, update it instead of starting over unless the user asks for a rebuild.
2. Extract context.
   • Product context: what it is, who it helps, outcomes, proof, pricing or delivery model if supplied.
   • Sector context: market, submarkets, geography, regulation, seasonality, and communities.
   • ICP: account types, roles, buying committee, users, qualification rules, and disqualification rules.
   • Competitors and alternatives: named competitors, incumbent workflows, internal substitutes, agencies, consultants, spreadsheets, manual process, or "do nothing" if supplied.
   • Pain points and triggers: urgent problems, trigger events, objections, adoption barriers, and timing signals.
   • Signal criteria: CRITICAL, HIGH, MEDIUM, and LOW definitions that map to the provided context.
   • Search seeds: keywords, phrases, queries, communities, hashtags, job titles, company types, events, and competitor terms.
   • Engagement constraints: approved positioning, banned claims, tone, channels, compliance notes, and approval requirements.
3. Normalize the output.
   • Convert vague inputs into concrete bullets without adding unsupported facts.
   • Keep labels stable so prospecting-signal-detector can scan the file quickly.
   • Separate facts from assumptions.
   • Mark inferred items as Inference: only when the inference is directly supported.
4. Validate before handoff.
   • Confirm every signal class has at least one criterion or is marked Needs input.
   • Confirm ICP includes both fit and disqualification guidance or marks missing parts.
   • Confirm search seeds are tied to context, not generic filler.
   • Confirm engagement constraints include "do not post without explicit user approval".
5. Handoff.
   • Tell the user where signal-context.md was written.
   • Suggest using prospecting-signal-detector with that file for scanning, reporting, or engagement.

Quality Rules

• Do not invent competitors, ICPs, sectors, signal criteria, proof points, or claims.
• Prefer exact phrases from source material for search seeds.
• Keep the file concise enough to load before every scan.
• Redact sensitive values in chat summaries, but preserve non-secret context in the file.
• If source files conflict, ask which source wins before overwriting important context.