Install
$ agentstack add skill-suraj-davariya-ai-job-search-job-application-assistant ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Purpose
This file is the Plane 1 knowledge anchor for the job application workflow (ARCH-0010). As a Plane 1 skill, it lives under .claude/skills/job-application-assistant/ and provides the AI assistant with structured knowledge — it does not execute binaries. When activated, this skill loads candidate identity and preference data from the companion profile files and applies the framework rules that govern the drafter-reviewer pipeline (ARCH-0030). The skill supports all four pipeline steps: fit evaluation, CV generation, cover letter generation, and interview preparation. All steps operate under the no-fabrication rule (ARCH-0007), the read-before-write invariant, and the writing-style rules in 03-writing-style.md.
Trigger Phrases
- "Apply to "
- "Evaluate this job posting"
- "Tailor my CV for "
- "Write a cover letter for "
- "Prepare me for an interview with "
- "Am I a good fit for ?"
- Any phrase about job applications, fit evaluation, or interview prep
Companion Files
| File | When to Read | |------|--------------| | 01-candidate-profile.md | Whenever candidate identity, education, experience, or skills data is needed | | 02-behavioral-profile.md | For behavioral/culture fit evaluation or when writing about personality fit | | 03-writing-style.md | Before generating any CV or cover letter content — rules always apply | | 04-job-evaluation.md | When evaluating fit for a job posting (Step 1 of pipeline) | | 05-cv-templates.md | When generating or tailoring a CV (Step 2 of pipeline) | | 06-cover-letter-templates.md | When generating a cover letter (Step 3 of pipeline) | | 07-interview-prep.md | When preparing for an interview or running practice sessions | | 08-legitimacy.md | Before drafting (Step 1) — assess posting legitimacy as a separate gate (REQ-8001) |
Sub-Commands (REQ-6003)
| Entry Phrase | Pipeline Step | Files to Read | |--------------|---------------|---------------| | "Evaluate this job posting" | Step 1 — Fit Evaluation | 04-job-evaluation.md, 08-legitimacy.md, 01-candidate-profile.md, 02-behavioral-profile.md | | "Write a CV for " | Step 2 — CV Generation | 05-cv-templates.md, 01-candidate-profile.md, 03-writing-style.md | | "Write a cover letter for " | Step 3 — Cover Letter | 06-cover-letter-templates.md, 01-candidate-profile.md, 02-behavioral-profile.md, 03-writing-style.md | | "Prepare for interview with " | Step 4 — Interview Prep | 07-interview-prep.md, 01-candidate-profile.md |
Missing context rule: Each sub-command can run without upstream steps having completed. If upstream context is missing, prompt: "I need [X] first — want me to run that step now?" Never silently fail or fabricate missing information (ARCH-0007). All sub-commands share 03-writing-style.md rules, the no-fabrication rule, and the verification checklist.
Contract
- Always read existing profile files before writing (read-before-write invariant)
- Respect
[UPPER_SNAKE_CASE]placeholder tokens — do not overwrite them with invented data - Default AI tool name in all generated documents: Claude Code (unless
[AI_TOOL_NAME]is set in01-candidate-profile.md— DEC-017) - Never fabricate skills, experience, or claims not present in the candidate profile (ARCH-0007)
- The full application pipeline (ARCH-0030) has 7 steps: 0=Parse, 1=Evaluate, 2=Draft, 3=Reviewer Dispatch, 4=Revise, 5=Compile+Verify, 6=Final Checklist. The Sub-Commands table above labels user-facing entry points (Fit Evaluation, CV Gen, Cover Letter, Interview Prep) using shorthand step names — these are not the same as the pipeline step numbers.
- The verification checklist (pipeline Step 6 — Final Checklist) runs exactly once at the end of the full
/applypipeline. It does NOT run during the reviewer pass (pipeline Step 3 — Reviewer Dispatch, a separate Claude Code invocation with fresh context per ARCH-0002). Never run the verification checklist inside the reviewer's context.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: suraj-davariya
- Source: suraj-davariya/ai-job-search
- License: MIT
- Homepage: https://suraj-davariya.github.io/ai-job-search/
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.