pegtalent-cli

Power-user terminal interface for PegTalent — the AI talent platform. Browse jobs, manage companies and teams, take real skill assessments with Judge0 grading, earn verifiable badges, and run recruiter workflows without leaving your terminal.

The package is pegtalent-cli; it installs a command named peg.

Install

npm install -g pegtalent-cli
peg --version

Requires Node.js ≥ 20.

First-time setup

The CLI talks to https://pegtalent.com by default — no configuration needed. Just authenticate with an API key (create one in the web app under Settings → API Keys):

peg login --token <your-api-key>
# or: peg auth login --token <your-api-key>

Self-hosting or local development? Point the CLI elsewhere:

peg config set apiEndpoint http://localhost:3000
# or, without persisting:  PEG_API_ENDPOINT=http://localhost:3000 peg ...
# or, per command:         peg --api-endpoint http://localhost:3000 ...

Check or change configuration any time:

peg config list
peg config set outputFormat json     # table | json | csv
peg config get apiEndpoint

Common commands

# Skill assessments (real, Judge0-graded)
peg assessment tracks                 # browse tracks + skills
peg assessment start <skill-id>       # start; downloads starter code to ~/.pegtalent/assessments
peg assessment status                 # current assessment + scores
peg assessment submit <file>          # grade your solution
peg assessment results [id]           # detailed result breakdown
peg assessment history                # earned badges

# Jobs & applications
peg job list
peg search "machine learning engineer"
peg apply <job-id>
peg badge list

# Profile
peg profile show
peg profile edit

# Employer / recruiter
peg company ...        peg job ...        peg candidate ...
peg screen ...         peg dashboard      peg analytics
peg team ...

# Agency
peg agency ...

# Fun
peg play              # PegJump terminal game

Global flags: --json, --csv, -q/--quiet, -v/--verbose, -y/--yes, --api-endpoint <url>. Run peg <command> --help for details on any command.

AI-assisted assessments — note

Some skills are AI-assisted (decomposition → coding → explanation). The CLI grades the code portion; the decomposition and explanation stages — which feed the AI-judged score dimensions — are completed in the web app. For those skills, finishing the full flow on the web yields a complete score. peg assessment start will flag when a skill is AI-assisted.

Updating

peg update                    # check whether a newer version is published
npm install -g pegtalent-cli  # upgrade

Development (monorepo)

# from repo root
npm run cli:build             # build apps/cli → dist
node apps/cli/dist/index.js   # run the local build

# from apps/cli
npm run dev                   # tsc --watch

License

MIT © PegTalent