--- name: pp-pbs description: "The only machine-readable Pakistani price panel — every urban centre and essential item the Bureau prices weekly, accrued locally because the Bureau itself does not keep the series. Trigger phrases: `pakistan spi`, `weekly sensitive price indicator`, `pakistani food prices by city`, `price of wheat flour in lahore`, `pakistan cpi city prices`, `cost of living pakistan by city`, `use pbs`, `run pbs`." author: "qazmataz" license: "Apache-2.0" argument-hint: " [args] | install cli|mcp" allowed-tools: "Read Bash" metadata: openclaw: requires: bins: - pbs-pp-cli install: - kind: go bins: [pbs-pp-cli] module: github.com/mvanhorn/printing-press-library/library/other/pbs/cmd/pbs-pp-cli --- # Pakistan Bureau of Statistics — Printing Press CLI ## Prerequisites: Install the CLI This skill drives the `pbs-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 `$HOME/.local/bin` on macOS/Linux and `%LOCALAPPDATA%\Programs\PrintingPress\bin` on Windows: ```bash npx -y @mvanhorn/printing-press-library install pbs --cli-only ``` 2. Verify: `pbs-pp-cli --version` 3. Ensure the reported install directory is on `$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 `$PATH` instead: ```bash go install github.com/mvanhorn/printing-press-library/library/other/pbs/cmd/pbs-pp-cli@latest ``` If `--version` reports "command not found" after install, the runtime cannot see the binary directory on `$PATH`. Do not proceed with skill commands until verification succeeds. Pakistan Bureau of Statistics publishes the richest sub-national price data in the country one machine-hostile weekly file at a time, indexes those files in a hand-edited JavaScript array, and deletes its own history: its live index reaches further back than the Internet Archive does, and one monthly annex is already served under two different months so one of them is lost. This CLI turns those scattered releases into one queryable panel and records every gap as a gap. Use spread and drift for questions the source cannot answer at all, coverage and verify to audit what you have, and revisions to catch what the Bureau quietly changed. ## When to Use This CLI Use this CLI for any question about Pakistani consumer prices that spans more than one release file: comparing an item across cities, following an item over weeks, building a custom weighted index, or auditing how complete a price series actually is. It is also the right tool when you need to prove what the Bureau published at a point in time, because it stores what the site later overwrites. Prefer it over reading the source files by hand whenever the answer needs more than one week. ## Anti-triggers Do not use this CLI for: - Do not use this CLI to fill in a missing week or estimate a price the Bureau did not publish; it reports gaps and will not interpolate. - Do not use it for a provincial or regional average; the Bureau does not publish the city weight vector, so any such figure would look official and be wrong. - Do not use it for Pakistani price history before July 2023; the older base series is no longer served and returns HTTP 404. - Do not use it for national headline inflation commentary alone, where the Bureau's own published index is the citable figure. - Do not use it for wage, fertilizer, cement or CNG series in this version; those sub-tables cover a different city set and are not yet parsed. ## Unique Capabilities These capabilities aren't available in any other tool for this API. ### Panel joins the source cannot answer - **`spread`** — See the cheapest and dearest city for any essential item in any week, and how that gap has moved over time. _Reach for this when the question is about geography rather than time — which city is an outlier on this item right now._ ```bash pbs-pp-cli spread "Onions" --as-of 2026-09-03 --agent ``` - **`drift`** — Follow one item's price across named cities and weeks, with missing weeks shown as gaps rather than filled. _Reach for this when the question is about time for a known item and known cities._ ```bash pbs-pp-cli drift "Wheat Flour Bag" --city lahore --city quetta --agent ``` - **`basket`** — Build your own price index over any subset of items using the weight vector PBS republishes every week. _Reach for this to build a food-only or fuel-only index the Bureau never computes._ ```bash pbs-pp-cli basket --item "Wheat Flour Bag" --item "Sugar Refined" --weight lowest --agent ``` - **`movers`** — Rank every item in a release by percent change recomputed from price levels, nationally or per city, with outliers flagged. _Reach for this when the question is which items moved this week, rather than one known item._ ```bash pbs-pp-cli movers --as-of 2026-09-03 --top 10 --flag-outliers --agent ``` ### Only the accrued store knows - **`weights`** — Read the item weight vector for any release, diff two releases, and check the totals sum to 100. _Reach for this before trusting a long index series — a silent re-weighting breaks comparability._ ```bash pbs-pp-cli weights --check-total --agent ``` - **`revisions`** — Detect releases the Bureau has rewritten, re-pointed, or deleted since you captured them. _Reach for this to prove your local copy is the surviving record, and to catch a silent restatement._ ```bash pbs-pp-cli revisions --all --agent ``` ### Honesty instrumentation - **`verify`** — Audit parsed cell values against the source's own arithmetic invariants, including the weight totals and min-avg-max ordering. _Reach for this before quoting any figure — it distinguishes a real price move from a parse bug._ ```bash pbs-pp-cli verify --all --agent ``` - **`coverage`** — Show which releases exist upstream, which the local store holds, and every hole named as a hole. _Reach for this whenever a result looks clean — a silently short series is the classic false positive._ ```bash pbs-pp-cli coverage --gaps --agent ``` ## Command Reference **file** — Static annexure/report files. All 202 live in one flat directory; the filename is the only variable. - `pbs-pp-cli file ` — Fetch one annexure, report, or CPI annex file by exact filename. Formats: . **index** — The PBS price-statistics release index. Carries TWO JavaScript literal arrays in the served HTML: `data` (148 weekly SPI releases, 2023-07-13..2026-09-03) and `cpidata1` (50 CPI months, July 2022..August 2026). - `pbs-pp-cli index` — Fetch the release-index page. ### Finding the right command When you know what you want to do but not which command does it, ask the CLI directly: ```bash pbs-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 ### Which city is gouging on onions this week ```bash pbs-pp-cli spread "Onions" --as-of 2026-09-03 --agent --select summary.min_city,summary.max_city,summary.range,summary.contributing_cities ``` Narrows a wide cross-city payload to just the extremes and the count of cities that actually reported, so an agent does not spend context on all the per-city rows. ### Food inflation for the poorest quintile ```bash pbs-pp-cli basket --item "Wheat Flour Bag" --item "Sugar Refined" --item "Pulse Gram" --weight lowest --json ``` Rebuilds an index over a chosen food subset using the Bureau's own lowest-quintile weights, which it republishes every week but never applies to a custom basket. ### Audit a series before quoting it ```bash pbs-pp-cli coverage --gaps --state-census --json ``` Prints missing releases and the per-release census of zero, blank and N.A. cells, so a clean-looking result can be checked for a silently short sample. ### Catch what the Bureau changed under you ```bash pbs-pp-cli revisions --all --json ``` Compares stored file hashes and index rows against the live site to surface rewritten bytes, re-pointed filenames, and releases that now return 404. ### Track how the basket itself was re-weighted ```bash pbs-pp-cli weights --diff 2023-07-13..2026-09-03 --json ``` The weight vector ships in every release with no change log upstream, so this diff exists only because the local store kept both editions. ## Auth Setup No authentication required. Run `pbs-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 pbs-pp-cli file mock-value --agent ``` - **Previewable** — `--dry-run` shows the request without sending - **Offline-friendly** — sync/search commands can use the local SQLite store when available - **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 ### Response envelope Commands that read from the local store or the API wrap output in a provenance envelope: ```json { "meta": {"source": "live" | "local", "synced_at": "...", "reason": "..."}, "results": } ``` Parse `.results` for data and `.meta.source` to know whether it's live or local. A human-readable `N results (live)` summary is printed to stderr only when stdout is a terminal AND no machine-format flag (`--json`, `--csv`, `--compact`, `--quiet`, `--plain`, `--select`) is set — piped/agent consumers and explicit-format runs get pure JSON on stdout. ## Paths and state Agents should treat the CLI's path resolver as part of the runtime contract: - Use `--home ` for one invocation, or set `PBS_HOME=` to relocate all four path kinds under one root. - Use per-kind env vars only when a specific kind must diverge: `PBS_CONFIG_DIR`, `PBS_DATA_DIR`, `PBS_STATE_DIR`, `PBS_CACHE_DIR`. - Resolution order is per-kind env var, `--home`, `PBS_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 `credentials.toml`, `data.db`, cookies, and auth sidecars. `state` contains persisted queries, jobs, and `teach.log`. `cache` contains regenerable HTTP/cache files. - Stored secrets live in `credentials.toml` under the data dir. Existing legacy `config.toml` secrets are read for compatibility and leave `config.toml` on the first auth write. - Run `pbs-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": { "pbs": { "command": "pbs-pp-mcp", "env": { "PBS_HOME": "/srv/pbs" } } } } ``` Fleet precedence: an inherited per-kind env var overrides an explicit `--home` for that kind. Use `PBS_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 `PBS_HOME`, or `doctor` will not find credentials 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, run: ```bash pbs-pp-cli recall "" --agent ``` 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": ["", "pbs-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 `pbs-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; `pbs-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 fold into one family so near-duplicate teaches do not fragment the store. Fire it after assembling your user-facing response