Skip to content
Back to search

company-career-prep

Evidence-led AI career skill: company research, targeted resumes, proof of work, interview practice and outcome tracking

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/om-umrania--company-career-prep/badge.svg)](/listings/om-umrania--company-career-prep)

HTML

<a href="/listings/om-umrania--company-career-prep"><img src="/api/listings/om-umrania--company-career-prep/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Requirements: Git and an agent that can read Markdown skills. Python 3 is only needed for repository checks. Live research needs the agent's web/search tools; native resume edits need an appropriate document editor and PDF rendering capability. No API keys or paid service are bundled.

Install in Codex

Clone the public repository, then copy the **complete nested skill folder**. The destination check preserves an existing installation.

git clone https://github.com/om-umrania/company-career-prep.git
cd company-career-prep
skill_root="${CODEX_HOME:-$HOME/.codex}/skills"
mkdir -p "$skill_root"
if [ -e "$skill_root/company-career-prep" ]; then
  echo "Existing skill found: compare and back it up before updating."
else
  cp -R skills/company-career-prep "$skill_root/"
fi

Start a new agent session if the installed skill is not yet listed, then use:

Use $company-career-prep for [company] and [role].
My target outcome is [outcome] by [date].
Use this JD: [URL or file]. Focus on [exact responsibility].
My resume and evidence are at [private paths].
Research current company direction and relevant industry work.
Produce a role-to-evidence map, tailored resume plan and next actions.

For another agent, point it at `skills/company-career-prep/SKILL.md` and ask it to follow the linked references. `$skill-name` discovery is host-specific; this repository does not install integrations for other products.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 126 lines
  1. # Company Career Prep
  2. [![Validate package](https://github.com/om-umrania/company-career-prep/actions/workflows/validate.yml/badge.svg)](https://github.com/om-umrania/company-career-prep/actions/workflows/validate.yml)
  3. [![License: MIT](https://img.shields.io/badge/License-MIT-teal.svg)](LICENSE)
  4. **Turn scattered career research into evidence, useful work and measurable next steps.**
  5. ![Career Evidence Loop: research, evidence, artifacts, action and learning](assets/career-evidence-loop.png)
  6. An agent skill for company research, job-specific resumes, proof of work and interview preparation—connected to an outcome ledger. Built for candidates navigating ambiguous roles, changing companies and incomplete information.
  7. [Quick start](#quick-start) · [Complete guide](docs/GUIDE.md) · [Methodology](docs/METHODOLOGY.md) · [Worked example](examples/fictional-product-role.md) · [Prompt library](docs/PROMPTS.md) · [Contributing](CONTRIBUTING.md)
  8. > This is an instruction package for an AI agent, not a hosted application, job board or autonomous application bot. It does not guarantee interviews or offers. Public examples are fictional; candidate data stays private.
  9. ## The problem it solves
  10. A job description, a founder's post and a resume often live in separate conversations. This skill connects them to the same decision: **what credible evidence should I show for this role, and what should I do next?**
  11. | Starting point | What the skill does | Reviewable output |
  12. |---|---|---|
  13. | A vague opportunity | Resolve the exact JD, eligibility and focus | Role brief with explicit unknowns |
  14. | Dozens of links and posts | Check dates, authors, relevance and access depth | Source register and material signals |
  15. | A generic resume | Map role needs to actual work and candidate evidence | Versioned resume and change rationale |
  16. | An experience gap | Define a bounded, relevant demonstration | Work-sample brief and validation plan |
  17. | Rehearsed but fragile answers | Elicit an attempt, challenge it, retry | Story bank and practice record |
  18. | Lots of activity, unclear progress | Track receipts, responses and stage transitions | Outcome ledger and next decision |
  19. ## Quick start
  20. Requirements: Git and an agent that can read Markdown skills. Python 3 is only needed for repository checks. Live research needs the agent's web/search tools; native resume edits need an appropriate document editor and PDF rendering capability. No API keys or paid service are bundled.
  21. ### Install in Codex
  22. Clone the public repository, then copy the **complete nested skill folder**. The destination check preserves an existing installation.
  23. ```bash
  24. git clone https://github.com/om-umrania/company-career-prep.git
  25. cd company-career-prep
  26. skill_root="${CODEX_HOME:-$HOME/.codex}/skills"
  27. mkdir -p "$skill_root"
  28. if [ -e "$skill_root/company-career-prep" ]; then
  29. echo "Existing skill found: compare and back it up before updating."
  30. else
  31. cp -R skills/company-career-prep "$skill_root/"
  32. fi
  33. ```
  34. Start a new agent session if the installed skill is not yet listed, then use:
  35. ```text
  36. Use $company-career-prep for [company] and [role].
  37. My target outcome is [outcome] by [date].
  38. Use this JD: [URL or file]. Focus on [exact responsibility].
  39. My resume and evidence are at [private paths].
  40. Research current company direction and relevant industry work.
  41. Produce a role-to-evidence map, tailored resume plan and next actions.
  42. ```
  43. For another agent, point it at `skills/company-career-prep/SKILL.md` and ask it to follow the linked references. `$skill-name` discovery is host-specific; this repository does not install integrations for other products.
  44. ## How the loop works
  45. ```mermaid
  46. flowchart LR
  47. A[Define outcome and exact role] --> B[Research current signals]
  48. B --> C[Map work to evidence]
  49. C --> D[Create and verify artifacts]
  50. D --> E[Take authorized action]
  51. E --> F[Observe actual results]
  52. F --> A
  53. C -. Missing proof .-> G[Bounded work sample]
  54. G --> D
  55. ```
  56. Research expands understanding. A focus decision narrows it. Artifacts make the claim inspectable. Feedback changes the next cycle. You can enter at any stage without running the entire loop.
  57. ## Choose your mode
  58. | Need | Ask for | Scope boundary |
  59. |---|---|---|
  60. | “Is this opportunity worth pursuing?” | Opportunity and evidence brief | No invented eligibility or fit score |
  61. | “Tailor this resume to one JD section” | Focused role version | Retain other material requirements |
  62. | “What changed at this company?” | Research refresh | Report accessible coverage, not “all posts” |
  63. | “I lack relevant proof” | Small work sample | Label proposed work and prototypes honestly |
  64. | “Help me practise” | Coached interview or discussion | Your attempt comes before model answers |
  65. | “Build the whole system” | Full career cycle | Requested local artifacts plus tracked next actions |
  66. ## What makes it evidence-led
  67. The skill separates official requirements, company claims, employee opinions, student analyses, candidate evidence and hypotheses. Recent LinkedIn activity can motivate a question; it cannot establish an official hiring requirement. A trend can change which real experience is emphasized; it cannot become experience the candidate never had.
  68. Every substantive resume version connects **JD clause → work process → company signal → candidate proof → editing decision**. Every action records an artifact version and an observed result. Reading volume and application volume are not treated as hiring success.
  69. ## Built for unstructured situations
  70. The workflow adapts Double Diamond, Jobs to Be Done, hypothesis-driven experimentation and Cynefin-informed probing. It uses different methods for different uncertainties instead of forcing every career question into one framework. [See sources, decision rules and limitations](docs/METHODOLOGY.md).
  71. ## Documentation map
  72. | Resource | Read it when |
  73. |---|---|
  74. | [End-to-end guide](docs/GUIDE.md) | You want inputs, stages, outputs and completion checks |
  75. | [Methodology](docs/METHODOLOGY.md) | You want to understand why each step exists |
  76. | [Prompt library](docs/PROMPTS.md) | You want copyable starting requests |
  77. | [Fictional walkthrough](examples/fictional-product-role.md) | You want to see the evidence chain in practice |
  78. | [Company brief template](skills/company-career-prep/references/company-brief.md) | You are starting a company/role workspace |
  79. | [Research protocol](skills/company-career-prep/references/research-to-results.md) | You need freshness, source quality and outcome rules |
  80. | [Privacy and limitations](docs/PRIVACY.md) | You are preparing material for sharing |
  81. | [Architecture](docs/ARCHITECTURE.md) | You are extending or integrating the skill |
  82. ## Validation and project status
  83. ```bash
  84. ./scripts/check.sh
  85. ```
  86. Checks cover package references, local documentation links, required assets, private-path leakage and skill metadata. They do not prove research accuracy, agent compliance or better hiring outcomes. The repository also includes a manual behavioral evaluation guide.
  87. Version 0.1.0: public documentation and reusable workflow. No outcome-uplift study, live LinkedIn integration, hosted dashboard or scheduled monitoring is claimed.
  88. ## Contribute
  89. Bring a reproducible failure, a redacted example or a focused improvement. See [CONTRIBUTING.md](CONTRIBUTING.md). Do not submit real resumes, private conversations or application records.
  90. Inspired by the workflow separation in [job-kit](https://github.com/yapsancode/job-kit) and the evidence-rich knowledge approach in [resume-builder](https://github.com/damienwin/resume-builder). This package uses original instructions; it does not bundle those applications. Framework attribution is in the methodology guide. License: [MIT](LICENSE).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.