feat(monitor+scoring): live dashboard counters + store discarded jobs#84
Merged
Conversation
Two bugs found during the first full pipeline run after the search milestone: 1. generate_queries self-loop: _needs_generate_queries checked state["raw_queries"] but the node writes state["queries"], so the router always saw an empty list and looped. Replaced the conditional self-edge with a direct edge to search_jobs. Also fixed the cache-hit path which read from state["raw_queries"] (always []) instead of the queries file. 2. anthropic_web search returned nothing: allow_tools: false in config was applied globally, including to the search LLM. The Claude CLI needs --dangerously-skip-permissions to invoke web-search tools. Factory now overrides allow_tools=True when task="search". Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Drop the redundant "→" link column; the run ID cell now carries the href so the table is one column narrower and every run is still one click away. Fixed the test that checked for the old trailing link cell pattern. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
_node_row_html computed total_tokens = in + out, dropping cache_read and cache_creation. This caused the pipeline execution table to show lower numbers than the grand total, making the two sections appear inconsistent. Now matches _usage_row_html which already counted all four buckets. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
build_llm(cfg["llm"]) without a task arg defaults to task="default" which resolves allow_tools=False. Company searches with url: hints invoke the Claude CLI and need --dangerously-skip-permissions to browse the web — same fix as search_jobs already had. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Grand total line now shows '≈Xk effective compute' in green when cache tokens are present (formula: new_in + out + 0.1×cache_read), making it clear that high cache-read is efficient, not wasteful. Per-node pipeline table replaces the single token total with 'Xin / Yout' and adds '/ Zcached' (green) when cache-read tokens are present for that node. Same change applied to the live-page JS so the live view is consistent. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…ctive Three connected fixes to improve search volume and quality: 1. url_validator: drop known aggregator/listing-page URL patterns before hitting Tavily (builtin.com, hnhiring, arc.dev listing pages, etc.). These pass URL validation because Tavily can fetch them, but they're search-result category pages, not individual postings — scoring rejects them at near-100% rate, wasting Tavily extract quota. 2. search_jobs: raise _DIRECTIVE_TARGET 30→50 and _DIRECTIVE_LLM_MAX 50→80 to target 30-50 validated individual postings per run. 3. web_search SEARCH_DIRECTIVE: instruct the LLM to search each major job board (WTTJ, LinkedIn /jobs/view, Lever, Greenhouse, Ashby, Workday) with dedicated queries rather than relying on broad web results. Explicitly forbid listing/search pages in the FORBIDDEN block so the LLM understands what counts as an individual posting. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Prompts extracted from web_search.py now live in: query/SEARCH_DIRECTIVE_PROMPT.md — directive (jobs search, URL candidates) query/SEARCH_COMPANY_PROMPT.md — company single-query search Edit these files to tune search behaviour without touching Python code. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
… fallback Primary path uses linkedin-api (unofficial voyager API, email/password auth). Falls back transparently to stickerdaniel/linkedin-mcp-server (Patchright browser) on any primary failure. MCP server installed under mcp_servers/ and ignored by git. - providers/search/connectors/linkedin.py: full implementation replacing stub - requirements.txt: add linkedin-api>=2.3.1, mcp>=1.0.0 - config/config.yaml: enable linkedin connector, max_concurrent: 1 - agent/nodes/search_jobs.py: add linkedin to _DEFAULT_MAX_CONCURRENT - tests/test_linkedin_connector.py: 19 unit tests, all passing - .gitignore: exclude mcp_servers/ (third-party local installs) One-time setup required: cd mcp_servers/linkedin-mcp-server && uv run -m linkedin_mcp_server --login Secrets needed in Infisical: LINKEDIN_EMAIL, LINKEDIN_PASSWORD Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
mypy cannot infer the Linkedin type from a lazy import inside a method, so it types self._client as None and flags .search_jobs() as attr-defined. The ignore is safe: the assignment on the line above guarantees non-None. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Live dashboard: - Duration counter increments live via JS requestAnimationFrame loop - Running node arrow (⟳) now rotates with CSS spin animation - Per-node time column counts live while node is active, shows final on complete - Jobs treated column (search_jobs, search_companies, analyze_jobs, store_results) - Discarded jobs section shows below stored jobs with real scores and reasoning Scoring pipeline: - LLM now scores ALL jobs instead of omitting low scorers - score_jobs_batch() returns (passed, discarded) tuple - discarded_jobs added to AgentState and live snapshots - analyze_jobs writes query/jobs_discarded.jsonl per run - store_results appends new discarded jobs to .data/discarded_jobs.jsonl (deduped by URL) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
bayrem
previously approved these changes
Jun 2, 2026
…column Both conflicts were main reverting the pipeline table to the old 5-column format (no Jobs column). Kept the HEAD version with the Jobs column and jobs_str logic intact. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
bjridicodes
dismissed
bayrem’s stale review
The merge-base changed after approval.
bayrem
previously approved these changes
Jun 2, 2026
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| for line in f: | ||
| try: | ||
| existing_urls.add(json.loads(line).get("url", "")) | ||
| except json.JSONDecodeError: |
score_jobs_batch return type changed to tuple in the scoring refactor; scoring_baseline.py was still treating the result as a plain list. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
bayrem
approved these changes
Jun 2, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
score_jobs_batchreturns(passed, discarded)tuple, discarded jobs stored to.data/discarded_jobs.jsonland shown in dashboard with real scores and reasoningLive dashboard changes
requestAnimationFrameusingrun_start_timefrom state@keyframes spinapplied to⟳on running nodenode_start_timesfrom graph; shows final time on completesearch_jobs,search_companies,analyze_jobs,store_resultsTest plan
pytest tests/— 57 tests passinfisical run --env=dev -- .venv/bin/python run.pyand openhttp://127.0.0.1:8765to confirm live counters workquery/jobs_discarded.jsonland.data/discarded_jobs.jsonlexist with scored entries🤖 Generated with Claude Code