--- name: pp-eu-tenders description: "The EU public procurement corpus searchable offline, with contactable leads from contract award winners. Trigger phrases: `find EU procurement contracts`, `who won construction contracts in Germany`, `leads from tender award winners`, `open tenders expiring soon`, `use eu-tenders`, `run eu-tenders-pp-cli`." author: "Mathias Michel" license: "Apache-2.0" argument-hint: " [args] | install cli|mcp" allowed-tools: "Read Bash" metadata: openclaw: requires: bins: - eu-tenders-pp-cli install: - kind: go bins: [eu-tenders-pp-cli] module: github.com/mvanhorn/printing-press-library/library/sales-and-crm/eu-tenders/cmd/eu-tenders-pp-cli --- # EU Tenders — Printing Press CLI ## Prerequisites: Install the CLI This skill drives the `eu-tenders-pp-cli` binary. **You must verify the CLI is installed before invoking any command from this skill.** If it is missing, install it first: 1. Install via the Printing Press installer. It defaults binaries to `$HOME/.local/bin` on macOS/Linux and `%LOCALAPPDATA%\Programs\PrintingPress\bin` on Windows: ```bash npx -y @mvanhorn/printing-press-library install eu-tenders --cli-only ``` 2. Verify: `eu-tenders-pp-cli --version` 3. Ensure the reported install directory is on `$PATH` for the agent/runtime that will invoke this skill. If the `npx` install fails (no Node, offline, etc.), fall back to a direct Go install (requires Go 1.26.6 or newer). This installs into `$GOPATH/bin` (default `$HOME/go/bin`), so add that directory to `$PATH` instead: ```bash go install github.com/mvanhorn/printing-press-library/library/sales-and-crm/eu-tenders/cmd/eu-tenders-pp-cli@latest ``` If `--version` reports "command not found" after install, the runtime cannot see the binary directory on `$PATH`. Do not proceed with skill commands until verification succeeds. Sync TED notices into a local SQLite store and turn award notices into outreach lists with leads, including email, phone and VAT id per winner when TED reports them. Rank open calls with score and deadline-heat, and analyze markets with concentration, buyer, winner, velocity and cpv-drift. No API key needed. ## When to Use This CLI Use this CLI to find EU public tenders, analyze who wins public contracts, or build B2B lead lists from contract award winners. It suits recurring sales prospecting, bid triage and procurement market research over TED data. ## Anti-triggers Do not use this CLI for: - Submitting or publishing notices to TED as an eSender - National below-threshold tenders that are not published on TED - Legal advice on procurement procedures ## Unique Capabilities These capabilities aren't available in any other tool for this API. ### Lead generation from award winners - **`leads`** — Get a contactable outreach list of companies that just won construction contracts: one row per winner with email, phone, city and VAT/HRB id when TED reports them, plus contract value and what is being built. _Reach for this when an agent needs sales leads from public contract awards instead of raw notices._ ```bash eu-tenders-pp-cli leads --country DEU --cpv 45 --days 30 --region DE2 --agent ``` - **`winner`** — Profile one company that wins public contracts: win timeline, total value, buyers, regions and latest contact data. _Use it before calling a lead or sizing up a competitor._ ```bash eu-tenders-pp-cli winner "Johann Bunte" --country DEU --agent ``` ### Bid prioritization - **`score`** — Rank open tenders by deadline urgency, contract value and keyword fit into a prioritized bid shortlist. _Use it when asked which open tenders are worth bidding on this week._ ```bash eu-tenders-pp-cli score --country DEU --cpv 45 --keywords Hochbau,Neubau --agent ``` - **`incumbents`** — For one open tender, see who previously won contracts from the same buyer in the same category. _Use it to judge your odds against the incumbent before writing a proposal._ ```bash eu-tenders-pp-cli incumbents 679227-2026 --agent ``` - **`deadline-heat`** — A ranked list of tenders closing within days, weighted by urgency, value and how few companies usually win this buyer's awards. _Use it for last-minute bid triage._ ```bash eu-tenders-pp-cli deadline-heat --country DEU --cpv 45 --days 14 --agent ``` ### Market and buyer intelligence - **`buyer`** — Profile a contracting authority: publishing cadence, CPV mix, typical contract values and repeat winners. _Use it before bidding to one authority or when investigating its award history._ ```bash eu-tenders-pp-cli buyer --name "Stadt München" --show-winners --agent ``` - **`concentration`** — See which companies capture what share of awarded value in a sector and country, with an HHI concentration score. _Use it to answer who dominates a procurement market._ ```bash eu-tenders-pp-cli concentration --country POL --cpv 4523 --top 5 --agent ``` - **`win-rate`** — See per buyer how many calls for tender end in published awards and how many distinct companies win. _Use it to compare buyers by openness before targeting them._ ```bash eu-tenders-pp-cli win-rate --country DEU --cpv 45 --min-calls 3 --agent ``` - **`dark-buyers`** — Flag contracting authorities whose tenders rarely produce public awards or keep going to one company. _Use it for procurement integrity screening; results are heuristic signals, not findings._ ```bash eu-tenders-pp-cli dark-buyers --country POL --cpv 45 --agent ``` ### Trends - **`velocity`** — See whether a procurement market is heating up or cooling off with weekly notice counts versus the same window last year. _Use it to spot demand shifts in a sector._ ```bash eu-tenders-pp-cli velocity --country DEU --cpv 45 --window 90d --compare 1y --agent ``` - **`cpv-drift`** — See which procurement categories grow or shrink year over year in a country's spending mix. _Use it for market sizing and policy research._ ```bash eu-tenders-pp-cli cpv-drift --country FRA --metric value --top 20 --agent ``` ## Command Reference **Search and lookup (live TED API)** - `eu-tenders-pp-cli notices --query "buyer-country=DEU AND classification-cpv=45000000"` — Expert-query search over TED notices (`--fields`, `--limit`, `--page`, `--pagination-mode`, `--scope`) - `eu-tenders-pp-cli notices get 680471-2026` — One notice by publication number, with winners and their contacts - `eu-tenders-pp-cli awards --country DEU --cpv 45` — Contract award notices with winner, buyer and value (`--winner`, `--year`, `--since`) - `eu-tenders-pp-cli deadline --country DEU --days 14` — Open calls for tender whose submission deadline is within N days - `eu-tenders-pp-cli cpv get 45500000` — CPV code lookup by code - `eu-tenders-pp-cli cpv search` (keyword argument, e.g. machinery) — CPV code lookup by keyword - `eu-tenders-pp-cli fields --contains winner` — TED expert-query field names **Local store (offline)** - `eu-tenders-pp-cli sync --country DEU --cpv 45 --since 90d` — Sync calls for tender and awards into SQLite (tables `notices`, `notice_winners`, `notices_fts`; `--type`, `--param key=value`, `--until`, `--limit`, `--max-pages`, `--db`) - `eu-tenders-pp-cli search "Neubau Schule"` — Full-text search over synced titles, buyers and winners - `eu-tenders-pp-cli sql "SELECT buyer_country, COUNT(*) FROM notices GROUP BY 1"` — Read-only SELECT against the store ### Finding the right command When you know what you want to do but not which command does it, ask the CLI directly: ```bash eu-tenders-pp-cli which "" ``` `which` resolves a natural-language capability query to the best matching command from this CLI's curated feature index. Exit code `0` means at least one match; exit code `2` means no confident match — fall back to `--help` or use a narrower query. `--json` (and other machine formats) keep that exit-2 contract and write `{"matches":[]}` on stdout so agents can inspect the envelope without treating a miss as success. ## Recipes ### Weekly construction-award leads ```bash eu-tenders-pp-cli leads --country DEU --cpv 45 --days 7 --new-only --agent --select winner_name,winner_email,winner_phone,winner_city,contract_value,title ``` Only companies not returned by earlier --new-only runs; this run's companies are recorded in the local lead_seen table. ### Group leads by company ```bash eu-tenders-pp-cli leads --country DEU --cpv 45 --days 90 --group-by company --agent ``` One row per company with win count and total value, ordered by biggest winners. ### Who dominates road construction ```bash eu-tenders-pp-cli concentration --country POL --cpv 4523 --top 10 --agent ``` Winner share and HHI for a market slice. ### Check the incumbent before bidding ```bash eu-tenders-pp-cli incumbents 679227-2026 --agent ``` Prior winners from the same buyer and category. ## Auth Setup No authentication required. Run `eu-tenders-pp-cli doctor` to verify setup. ## Agent Mode Add `--agent` to any command. Expands to: `--json --compact --no-input --no-color`. Global format flags share one contract on promoted, novel, sync, and `--deliver` paths: - `--json` — one JSON document on stdout (sync progress events go to stderr) - `--compact` — keep identity/status/timestamp fields; does not change the document vs stream shape - `--csv` / `--plain` — tabular rows (collection envelopes unwrap to the row array) - `--quiet` — one identity value per row, no envelope - **Pipeable** — JSON on stdout, errors on stderr - **Filterable** — `--select` keeps a subset of fields. Dotted paths descend into nested structures; arrays traverse element-wise. Critical for keeping context small on verbose APIs: ```bash eu-tenders-pp-cli leads --country DEU --cpv 45 --agent --select winner_name,winner_email,contract_value ``` - **Previewable** — `--dry-run` shows the request without sending - **Non-interactive** — never prompts, every input is a flag - **Read-only** — do not use this CLI for create, update, delete, publish, comment, upvote, invite, order, send, or other mutating requests ## Paths and state Agents should treat the CLI's path resolver as part of the runtime contract: - Use `--home ` for one invocation, or set `EU_TENDERS_HOME=` to relocate all four path kinds under one root. - Use per-kind env vars only when a specific kind must diverge: `EU_TENDERS_CONFIG_DIR`, `EU_TENDERS_DATA_DIR`, `EU_TENDERS_STATE_DIR`, `EU_TENDERS_CACHE_DIR`. - Resolution order is per-kind env var, `--home`, `EU_TENDERS_HOME`, XDG (`XDG_CONFIG_HOME`, `XDG_DATA_HOME`, `XDG_STATE_HOME`, `XDG_CACHE_HOME`), then platform defaults. - `config` contains settings like `config.toml` and profiles. `data` contains `data.db` (notices, notice_winners, notices_fts, lead_seen, learn-loop tables) and `feedback.jsonl`. `state` contains `teach.log`. `cache` contains regenerable HTTP cache files. - Run `eu-tenders-pp-cli doctor --fail-on warn` to surface path warnings. `agent-context` exposes a schema v4 `paths` block for agents that need the resolved dirs. - For MCP, pass relocation through the MCP host config. The MCP binary does not inherit CLI flags: ```json { "mcpServers": { "eu-tenders": { "command": "eu-tenders-pp-mcp", "env": { "EU_TENDERS_HOME": "/srv/eu-tenders" } } } } ``` Fleet precedence: an inherited per-kind env var overrides an explicit `--home` for that kind. Use `EU_TENDERS_HOME` or per-kind vars as durable fleet levers, and use `--home` only for a single invocation. Relocation is not reversible by unsetting env vars; move files manually before clearing `EU_TENDERS_HOME`, or `doctor` will not find the store and config left under the former root. ## Automatic learning This CLI ships a self-capturing learning loop. The CLI does its own bookkeeping: every invocation is journaled locally, a failed flag followed by a corrected retry auto-derives a `flag_alias` candidate, and a `teach` on a query family without a playbook auto-synthesizes a `playbook_candidate` from the session's journal. Your job is judgment only: `recall` first, act on surfaced candidates, `teach` the final answer, `playbook amend` when you observe a correction. You never record failures by hand. ### Step 1: `recall` before any discovery Before list/search/drill commands on a new user question, pass the question as an argv or MCP tool argument to `recall --agent`. Do not interpolate user-controlled text into a shell command line. Quoted `recall ""` breaks on an apostrophe, which is ordinary English. A quoted heredoc breaks when a body line equals the delimiter, and that delimiter is published in these docs. Write the question with a non-shell file-writing tool, then read it back as data: ```bash # Write the question verbatim with your file-writing tool (no shell involved). # Command substitution on a file only ever yields data — the shell never # parses the file's bytes as syntax. QUERY=$(cat /path/to/question.txt) eu-tenders-pp-cli recall "$QUERY" --agent ``` Prefer MCP: pass the question as the tool's query argument. `"$QUERY"` after a file read is argv-safe; putting the question itself in the command text is not. The response envelope: ```json { "query": "...", "normalized": "", "query_entities": ["..."], "found": true | false, "match_score": 0.0, "results": [ { "resource_id": "...", "resource_type": "...", "venue": "...", "confidence": 2, "entity_match": "exact|partial|unknown", "source": "taught|preseed|pattern", "warnings": ["..."] } ], "mismatches": [ /* only when --debug-mismatches */ ], "warnings": [ /* top-level */ ], "candidates": [ { "id": 12, "class": "flag_alias | playbook_candidate", "summary": "...", "sightings": 3, "last_seen": "...", "rationale": "...", "next_action": ["", "eu-tenders-pp-cli learnings confirm 12"] } ], "playbook": { "query_family": "...", "playbook": { "steps": [ { "cmd": "", "purpose": "..." } ], "entity_slots": ["$ENTITY"], "expected_tool_calls": 3 }, "slots_resolved": { "$ENTITY": { "token": "", "canonical": "" } }, "notes": "" }, "notes": "" } ``` Empty-store short-circuit: if the store has no learnings, playbooks, or candidates yet (recall finds nothing and `learnings list` and `learnings candidates` are both empty), skip recall for the rest of this session instead of taxing every query; resume recall-first once something has been taught. ### Step 2: decision tree Read `candidates`, `playbook`, `notes`, `results[0]`, and warnings in that order: ``` if Candidates present (warnings include "candidates_present"): -> candidates are try-then-confirm, never facts. Follow each candidate's two-step next_action verbatim: run the trial command first, then run `learnings confirm ` only after the trial verified the behavior. Reject a wrong candidate with `learnings reject `. -> NEVER re-teach something recall surfaced as a candidate; confirm or reject that candidate instead of teaching a duplicate. -> candidates ride alongside playbooks and resource hits, not instead of them; continue with the branches below after acting on them. if Playbook present: -> READ Playbook.notes verbatim FIRST (workarounds + gotchas the CLI surface doesn't expose) -> replay Playbook.steps in order, substituting Playbook.slots_resolved entries for the entity slot tokens. If a step's slot is unresolved, fall back to discovery for that step only. -> the Playbook's expected_tool_calls is a budget; if you find yourself running materially more, record the divergence via `eu-tenders-pp-cli playbook amend` at end-of-session. elif Notes present (no Playbook): -> read Notes verbatim before any discovery step; they carry known gotchas for this query family even when no structured choreography exists yet. elif Found AND Results[0].EntityMatch == "exact" AND Results[0].Confidence >= 2: -> skip discovery; fetch live data for Results[*].ResourceID in parallel elif Found AND Results[0].EntityMatch == "partial": -> candidate hint, NOT a hit; read the resource title to validate before trusting elif (any row in Mismatches[] when --debug-mismatches was passed): -> treat as cold start; the stored learning is for a different entity (different canonical resolved from query_entities) else: // Found == false, no playbook, no notes -> cold start; run discovery normally; teach the answer afterward (Step 4). If the family has no playbook yet, that teach auto-synthesizes a playbook candidate from this session's journal - you do not need to record one by hand. ``` Playbook and Notes are orthogonal to the per-resource path. A recall response can carry both a Playbook AND a `Results[]` hit - use both: the Playbook tells you which choreography to run; the resource hits short-circuit specific steps. Default to skipping `mismatches`; pass `--debug-mismatches` only when investigating cold-start surprises. Candidate judgment details: `learnings confirm ` prints the candidate's full payload before materializing it - check that the printed payload matches the behavior you verified. `learnings reject ` tombstones the derivation signature so the same candidate does not resurface. The envelope carries only the few candidates worth acting on now; `eu-tenders-pp-cli learnings candidates` lists the full open set. Graceful degradation: if `learnings confirm` is an unknown command, you are driving an older binary - ignore the candidates guidance and follow the rest of the protocol. ### Step 3: always read `warnings` - `low_confidence`: row exists at `confidence<2`. Treat as a hint, not a skip-discovery hit. - `resource_not_in_store`: the local store doesn't have the resource the learning points at. The match validator couldn't classify entities — direct-fetch and re-evaluate. - `cross_alias_match` (per-result): the row was taught under a different alias and matched the live query's canonical via `entity_lookups` (e.g., a "USA" teach satisfying a "United States" recall). Trust the resource_id. - `similar_shape_different_entity:` (top-level): a structurally matching row exists but its canonical entity differs from the live query's. Treated as cold start; the warning carries the conflicting canonical as a hint, but the row is NOT promoted into Results. - `ambiguous_alias` (top-level): a single query entity resolved to multiple canonicals (e.g., "Cards" → Arizona Cardinals + St. Louis Cardinals). Surface the ambiguity from context before committing to a resource. - `candidates_present` (top-level): the envelope carries a `candidates` section. Handle it via the candidates branch in Step 2 before anything else. - Top-level `no_learnings_for_query_family`: the table had no rows above the Jaccard floor. Pure cold start. ### Step 4: `teach &` after finalizing your response - always Teaching is unconditional. After resolving a query the store could not answer, background-teach the final resource mapping - no call-count threshold, no judging whether it was "worth" learning. The teach is the anchor of the loop: it triggers playbook synthesis for a family without a playbook, and same-referent phrasings fol