Skip to content

docs(readme): update for LinkedIn, discarded jobs, live monitor features #85

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
bjridicodes wants to merge 1 commit into
base: main
Choose a base branch
from
Open

docs(readme): update for LinkedIn, discarded jobs, live monitor features #85

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 20 additions & 5 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ A LangGraph-based agent that autonomously discovers, scores, and tracks job oppo

1. **Loads context** — reads your CV files (`query/resume/`), generates search queries deterministically from `config/search_config.yaml` (positions × locations cross-product), and loads target companies with their ATS hints
2. **Searches for jobs** — one directive LLM prompt returns job URLs only (no fabricated descriptions); Tavily extract validates each URL and pulls real posting content (hallucinated or unreachable URLs are dropped); company ATS boards (Greenhouse, Lever, Ashby) are queried via direct API — zero LLM tokens for ATS; all results deduplicated and checkpointed to `query/jobs_found.jsonl`
3. **Scores matches** — single LLM call scores all jobs against your CV; keeps only jobs above a configurable threshold
3. **Scores matches** — single LLM call scores all jobs against your CV; jobs above the threshold are stored for action; jobs below the threshold are kept with their real score and one-line reason in a separate discarded log for review
4. **Stores results** — deduplicates by content-hash and writes to local JSON and/or cloud storage (Google Drive, OneDrive, Dropbox)
5. **Notifies you** — sends a digest to Telegram, Slack, email, or WhatsApp

Expand Down Expand Up @@ -66,6 +66,10 @@ python3 -m venv .venv
# TAVILY_API_KEY — for URL validation and extraction (required)
# FRANCE_TRAVAIL_CLIENT_ID/SECRET — optional free job board API
# ADZUNA_APP_ID/KEY — optional free job board API
# LINKEDIN_EMAIL, LINKEDIN_PASSWORD — for LinkedIn connector (unofficial API)

# 2b. LinkedIn MCP fallback setup (one-time, optional)
# cd mcp_servers/linkedin-mcp-server && uv run -m linkedin_mcp_server --login

# 3. Add your CV
# Drop a PDF or .md file into query/resume/
Expand Down Expand Up @@ -100,9 +104,12 @@ search:
- name: anthropic_web # primary: LLM directive search → Tavily extract
max_results_per_query: 4
- name: france_travail # optional free API — francetravail.io
enabled: false
enabled: true
- name: adzuna # optional free API — developer.adzuna.com
enabled: false
enabled: true
- name: linkedin # unofficial API + MCP browser fallback
enabled: true # requires LINKEDIN_EMAIL / LINKEDIN_PASSWORD
max_concurrent: 1

storage:
provider: local # local | google_drive | onedrive | dropbox
Expand Down Expand Up @@ -164,7 +171,15 @@ When the pipeline is launched via Claude Code in VS Code (which blocks the TUI)
🌐 Live monitor: http://127.0.0.1:8765/ (run_id=abc12345)
```

The page polls `/state.json` every second, refreshes the pipeline table, token-spend block, and job cards in place, and stops polling automatically when the run finishes. The same HTML template is reused for the static post-run report at `logs/runs/run_*.html` (just without the JS poll block).
The page polls `/state.json` every second and provides live feedback:

- **Duration counter** — increments every frame via `requestAnimationFrame`
- **Spinning arrow** — the active node's `⟳` rotates with a CSS animation
- **Per-node timer** — counts up while the node runs, locks to the final time on completion
- **Jobs column** — shows how many jobs each key node has treated so far
- **Discarded jobs section** — below stored jobs, shows all sub-threshold results with their real LLM score and one-line rejection reason

The same HTML template is reused for the static post-run report at `logs/runs/run_*.html` (just without the JS poll block).

CLI flags (issue #62):

Expand All @@ -190,7 +205,7 @@ Per-model and per-node totals are stored on the final state as `token_usage` (sh
| Orchestration | LangGraph |
| LLM interface | LangChain (Anthropic Claude / OpenAI) |
| Search | Claude web search (directive prompt) + Tavily extract (validation + content) |
| Job boards | France Travail, Adzuna (optional) |
| Job boards | France Travail, Adzuna, LinkedIn (unofficial API + MCP browser fallback) |
| ATS boards | Greenhouse, Lever, Ashby (unauthenticated HTTP) |
| Terminal UI | Rich |
| Storage | Local JSON (Google Drive / OneDrive / Dropbox) |
Expand Down
Loading