Add an interview-prep skill for claude.ai chat that builds a stage prep pack from the career master, pitfalls, posting and submitted CV and letter, runs a mock interview and drafts thank-you and follow-up notes, ported from ai-job-search (MIT)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
b481cb5692
commit
5f673d0a3f
7 files changed
+372
No files matched your search
@@ -0,0 +1,61 @@
|
||||
# STAR answers and gap bridges
|
||||
|
||||
## STAR format
|
||||
|
||||
- **Situation**: the context and the problem, in two sentences.
|
||||
- **Task**: what the owner specifically was responsible for.
|
||||
- **Action**: what he did, with methods and tools. "I", not "we", unless the point is the team.
|
||||
- **Result**: the outcome, measured where the career master has the number; adoption or impact otherwise.
|
||||
- **Learned**: one sentence on what he learned or would do differently.
|
||||
|
||||
Length: 1-2 minutes spoken, roughly 150-300 words written. Specific beats complete; cut setup the question does not need.
|
||||
|
||||
Written in the pack as:
|
||||
|
||||
```
|
||||
### Migration to managed queues (ownership, technical judgement)
|
||||
S: ...
|
||||
T: ...
|
||||
A: ...
|
||||
R: ...
|
||||
Learned: ...
|
||||
Use for: "a project you owned end to end", "a hard technical decision"
|
||||
Source: career master, role 2, project "queue migration"
|
||||
```
|
||||
|
||||
## Rules
|
||||
|
||||
1. Every S, T, A and R element traces to the career master. Name the source line.
|
||||
2. Never add a number, tool, title, team size, duration or outcome the career master does not state. If a story would be stronger with a metric that is missing, ask the owner for it; if he gives one, it goes into "Facts to add to the career master".
|
||||
3. A story must not contradict the CV or letter (dates, titles, scope). Check against the pitfalls before presenting.
|
||||
4. Cover different competencies across the pack. Aim for 4-6 core stories that map to most questions, rather than one story per question.
|
||||
5. Pick stories that match the posting's competencies, and the angle of the interviewer who will hear them.
|
||||
|
||||
## When no story fits: the honest bridge
|
||||
|
||||
For a question or requirement the career master does not cover:
|
||||
|
||||
1. **Acknowledge** the gap plainly, in one sentence. No apology, no deflection.
|
||||
2. **Adjacent experience**: the closest real experience and what transfers, with a fact from the career master.
|
||||
3. **Learning path**: what the owner has done or will do to close it (course, side project, reading, a first-90-days plan). Only steps he has actually taken count as past tense.
|
||||
|
||||
Example (generic):
|
||||
|
||||
> I haven't run Kubernetes in production. I have run containerised services on a managed platform for three years, including the scaling and incident work that goes with it, so the concepts are familiar. I've been working through the CKA material and set up a test cluster for a side project; in the first months I'd expect to pair with whoever owns the platform.
|
||||
|
||||
Then offer a **STAR stub** for the owner to fill, in case he has a story the career master does not record:
|
||||
|
||||
```
|
||||
### [Working title] (competency: ...)
|
||||
S: [When and where? What was going on?]
|
||||
T: [What were you responsible for?]
|
||||
A: [What did you do, step by step? Which tools?]
|
||||
R: [What changed? Any number?]
|
||||
Learned: [?]
|
||||
```
|
||||
|
||||
A filled stub is a new fact: it goes into "Facts to add to the career master" before it is used.
|
||||
|
||||
## Calibration to the owner
|
||||
|
||||
Write in the owner's register as it shows in his cover letter and the answers he gives in the session: sentence length, formality, how much he hedges or quantifies. Coach toward his best version of his own voice, not a generic ideal.
|
||||
Reference in new issue
Block a user