| name | opportunity-report-next-steps-one |
| description | Produces next-steps markdown for a single opportunity from opportunities.yaml using opportunity_report_next_steps_one template. Use when the user asks for one-opportunity next actions, 24–72h moves, or snapshot for a given id. |
Report — next steps for one opportunity
Role: Assistant — career / opportunity tracking (see docs/AGENT_ROLES.md).
Context to load:
data/opportunities.yaml- Output skeleton:
templates/opportunity_report_next_steps_one.md
Scope:
- Load only specific YAML record.
- Skip unrelated opportunity/role reports unless user asks.
Task:
- Select record by
id: <OPPORTUNITY_ID> (or company + role — resolve ambiguity).
- Produce Snapshot + Next actions from
next_action, next_action_date, notes, stage.
- Populate Contacts table from
contacts[]; if empty: "None recorded."
- Populate Check here (manual) table from
links; if empty: "None recorded."
- Add Blockers / risks from notes and stage.
- Add Suggested 24–72h moves aligned with
docs/framework/workflows.md (Workflow 2).
- Assumptions / Risk.
Output: Markdown for data/reports/pipeline/opportunity-next-steps-<id>-<YYYY-MM-DD>.md or paste-only.
Constraints:
- Full detail stays under
data/ (gitignored — real names fine).
- No fabrication. Every sentence traceable to YAML field (
stage, next_action, notes, history, contacts, links). No career advice, invented timelines, assumed details.
- Missing data → state explicitly. No supporting data → write "No data recorded." No gap-filling.
- Assumptions section mandatory. Any interpretation beyond direct field read must be listed, labeled as inference.
My inputs (collect, confirm, or ask):
- If opportunity id (or company + role) not in memory/chat/
opportunities.yaml: ask before generating.
- If id inferred from company + role: state resolved id, ask user to confirm or correct before final output.
- Opportunity id (or company + role):
<...>
