Story Bank

Story Bank

A once-built, forever-refined asset. Behavioral questions are predictable: "tell me about a conflict", "a failure", "a time you led without authority". A good story bank means you never improvise. You reach for the right pre-built story and adapt it. Build it once, refine it after every interview.

When to use

• "Build my story bank", "prep my behavioral stories", "help me with STAR answers"
• Before the first real interview in a search, because interview-prep is far stronger when this exists
• Anytime a new strong story surfaces (a debrief, a finished project) and needs to be captured
• The orchestrator routes here when stories/story-bank.md is empty or thin

Inputs

• Reads: workspace/profile/profile.md and workspace/documents/master-resume.md (achievements are raw story material), plus any existing workspace/stories/story-bank.md
• Needs from user: the actual events. You cannot write these. You extract them by interview. The resume tells you WHERE the stories probably are ("led the replatform", "turned around churn"), and the user supplies what actually happened.

The competency map (target coverage)

A complete bank covers these ten areas. Most real interviews draw from this set. Aim for 8–12 stories. A single strong story often covers 2–3 competencies.

1. Leadership / influence without authority. Moved people/a decision when you didn't own them
2. Conflict / disagreement. Pushed back, disagreed with a boss or peer, navigated tension
3. Failure + lesson. Something that genuinely went wrong, and what you changed after
4. Ambiguity / uncertainty. Acted well without a clear spec or a known answer
5. Data-driven decision. Used evidence to pick a direction or kill something
6. Cross-functional collaboration. Shipped across eng/sales/design/finance boundaries
7. Scale / biggest impact. The largest thing you moved (revenue, users, headcount, scope)
8. Turnaround / fixing something broken. Inherited a mess and repaired it
9. Initiative / from scratch. Saw a gap, started something nobody asked you to
10. Prioritization under constraint. Too little time, budget, or people, and chose well and defended it

Workflow

1. Mine the resume first. Read profile.md and master-resume.md. For each substantial role or achievement, form a hypothesis: "this replatform is probably a leadership and ambiguity story, and this churn number is probably a turnaround story." Draft a candidate list of 10–15 story seeds mapped to competencies. This makes the interview targeted, not open-ended.
2. Interview to extract, batched first, then dig. Present the seeds and ask the user to confirm which are real and which are their strongest. Then for each chosen story, pull the STAR detail with these prompts (ask the ones that are missing, don't interrogate mechanically):
   • Situation: "Set the scene. Company, when, and what was at stake if it went wrong?"
   • Task: "What specifically was YOURS to own here? Not the team's. Yours."
   • Action: "Walk me through what you personally did. What were the 2–3 hardest decisions and the tradeoffs?" (This is the meat. Push past "we" to "I". Interviewers score the action, not the outcome.)
   • Result: "What changed because of you? Put a number on it: revenue, %, time, scale, retention. If there's no hard number, what is the concrete before and after?"
   • Reflection: "Knowing what you know now, what would you do differently?" (Every strong story needs an honest lesson. It's how you answer 'what would you change' and it makes the story credible.) Batch the opening confirmation in one message, then go story by story so the detail stays deep.
3. Interrogate the numbers. For each Result, ask where the number came from. Real metrics only. If the user isn't sure, mark it [metric TK, user to confirm] and keep it visible. Never round up and never invent a plausible figure.
4. Write each story to stories/story-bank.md using templates/story-template.md. Capture Situation, Task, Action, and Result, plus Tags (searchable keywords), Maps to competencies (which of the ten it proves), Strength (1–5, how compelling and well-evidenced), and 1–2 reusable soundbites, crisp one-liners you can drop into a cover letter, outreach note, or interview intro.
5. Rate strength honestly. A 5 has a specific situation, clear personal action, and a hard quantified result. A 2 is vague or has no outcome. Low-strength stories are fine to keep, but flag them. They signal where the user needs a better example.
6. Run the coverage check. Map every story to the ten competencies and list which areas have no strong (4–5) story. Report the gaps plainly: "You have nothing strong for conflict or failure, the two most-asked. Let's find one." Then prompt the user to fill them, or note them as known weak spots to shore up before interviews.
7. Confirm and hand off. Show the user the map and gaps. Point them to interview-prep, which will pull specific stories against a specific role and stage.

Output

• Writes to: workspace/stories/story-bank.md (append, never overwrite, since this file grows over the whole search)
• Format:
  • A short header: last-updated date + the competency-coverage map (which stories cover which of the ten areas, and the gaps)
  • Then one entry per story in templates/story-template.md format
  • Keep the strongest stories near the top

Rules & quality bar

• No fabrication. You extract stories, you never author them. If the user can't source a competency from real experience, it stays a gap. You do not invent a plausible one.
• Numbers are real or marked TK. No invented or rounded-up metrics. A visible [metric TK] beats a confident fake.
• Action is first-person. Interviewers score what the candidate did. Push every "we" toward "I did X". If the user truly only contributed a slice, the story says exactly that slice.
• Gaps stay visible. The coverage map's whole value is showing what's missing. Never paper over an empty competency to make the bank look complete.
• This asset is cumulative. Append across the search. interview-followup adds newly surfaced questions and better answers here after each round. Don't rebuild from scratch each session.
• 8–12 is the target, not a quota. Six genuinely strong stories beat twelve weak ones. Depth over count.

Related skills

• Reads from: job-search-setup (profile + master resume supply the story seeds)
• Feeds: interview-prep (maps stories to a role's likely questions), cover-letter (soundbites), networking-and-referrals and outreach (soundbites), interview-followup
• Fed by: interview-followup, which appends newly surfaced questions and stronger answers here after every debrief, so each interview improves the bank
• Routed here by: search-and-scale-os when stories/story-bank.md is empty or coverage is thin