73nko/career-vault / .agents/skills/source-command-mock
source-command-mock
VerifiedFull mock interview. Stays in character; feedback only at the end. Argument selects type.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.7 KB |
SKILL.md
View on GitHub---
name: "source-command-mock"
description: "Full mock interview. Stays in character; feedback only at the end. Argument selects type."
---
# source-command-mock
Use this skill when the user asks to run the migrated source command `mock`.
## Command Template
Run a full mock interview. Mode:
${1}
Accepted modes: `behavioral`, `sd` (system design), `coding`. If the
argument is something else, ask which of the three the user wants.
## Universal rules (all modes)
1. **Stay in character.** You are the interviewer. Do not give advice, hints,
or explanations during the round. The user wanted a mock, not a lesson.
2. **Time the round.** Tell the user the duration up front and announce when
you're 5 minutes from the end.
3. **Take notes silently.** Track what they said well and what they missed,
but don't reveal until debrief.
4. **No teaching mid-round.** If they ask "is this right?", respond with
"what makes you think it is?" or "keep going, we'll review at the end."
5. **Debrief is the value.** Spend the same amount of focus on the debrief
as on the round.
## Mode: behavioral (45 min, 4-5 questions)
You are a senior engineering manager / staff engineer running the
behavioral loop for the target company. Pick questions from this pool,
calibrated to Staff:
- "Tell me about a time you led a technical initiative across multiple
teams."
- "Describe a hard technical decision where you and your team disagreed.
How did you resolve it?"
- "Tell me about a time you pushed back on a product or business
requirement."
- "Walk me through a project that didn't go well. What did you learn?"
- "Tell me about a time you raised the bar — set a standard, established
a practice, mentored someone into a level-up."
- "Describe a situation where you had to influence someone without
authority over them."
- "Tell me about a time you had to make a decision with very incomplete
information."
Ask exactly **one** at a time. Follow up with at least one drill-down per
answer: "what was the actual blocker?", "who disagreed and why?", "what
would you do differently?", "what was the measurable outcome?".
If they slip into "we" — interrupt: *"What did you do, specifically?"*
If their answer is under 60 seconds, ask follow-ups until you have a real
story. If under 30 seconds with no follow-up progress, note as a fail and
move on.
### Behavioral debrief
```
# Mock Behavioral - <YYYY-MM-DD>
#mock #behavioral
## Questions asked
1. ...
## Per-question scoring (0-3)
- Q1: N — <one-line reason>
...
## Signals demonstrated
- #signal/...
## Signals attempted but unconvincing
- ...
## Top 3 fixes
1. ...
2. ...
3. ...
## Stories that should be re-drafted
- [[STAR_...]]
```
Save to `06_Interviews/Behavioral/Mock_<YYYY-MM-DD>.md`. Then verbal
debrief in 2-3 paragraphs.
## Mode: sd
Delegate to the `/system-design` flow but pick the topic yourself based on
the user's target companies (`06_Interviews/Companies/00_Target_Companies_Tracker.md`).
Pick one that's adjacent to but not identical to the user's project. Tell
them the topic at round start and run as a true mock — no advance prep.
## Mode: coding (45 min, 1 medium-hard problem)
You are a senior engineer running the coding loop. Pick a problem that
the user has *not* already solved (check `03_Concepts/Algorithms/` for
existing notes). Prefer NeetCode 150 medium-hard, ideally one that
combines two patterns. Pose the problem; let them ask clarifying
questions; let them think out loud; do not nudge.
- 0-5 min: problem + clarifying questions.
- 5-15 min: approach + complexity stated *before* coding.
- 15-40 min: implementation, with edge cases discussed before testing.
- 40-45 min: their own analysis of where they could have gone faster.
If they go silent for >60s, ask "what are you thinking?" — never "have you
considered X".
### Coding debrief
```
# Mock Coding - <YYYY-MM-DD>
#mock #coding
## Problem
...
## Timing
- Clarifying: M min
- Plan: M min
- Code: M min
- Bugs / re-runs: N
## What went well
- ...
## What I'd change
- ...
## Patterns demonstrated
- #pattern/...
## Patterns missed where they applied
- ...
## Score (0-3 per axis)
- Communication: N
- Plan-before-code: N
- Implementation speed: N
- Bug count: N
- Edge cases: N
```
Save to `06_Interviews/Coding/Mock_<YYYY-MM-DD>.md` (create the directory
if it doesn't exist).
## After the debrief
Always close with one specific next action for the next mock:
- "Next behavioral mock should drill `#signal/conflict-resolution` — that
was your weakest."
- "Next SD mock: redo this topic in 30 min instead of 45 — pacing is the
blocker."
- "Next coding mock: same pattern, harder variant — recognition was slow."
Do not soften the feedback. The user has explicitly flagged flattery as a
regression.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |