@pieai/pro-gov 0.7.3 → 0.9.0

package/assets/profiles/engineering-runtime/profile.md

@@ -38,7 +38,7 @@ content under the target project's ownership. Optional Lefthook and CI
 guardrails are also consumer-owned when installed; `doc-gov doctor` checks that
 their required governance commands remain present.
 
-For non-trivial engineering work, run `pro-gov learn recall --query "<task
+For non-trivial engineering work, run `pnpm pro-gov learn recall --query "<task
 summary>"` before changing files so existing `docs/reference/learnings/**` and
 `CONCEPTS.md` lessons can shape the work. After completed, verified, non-trivial
 work, let `capture-learning` make a visible soft closeout decision. Capture only

package/assets/public-agent-assets/bundles/base-governance.json

@@ -5,7 +5,6 @@
   "assets": [
     "pie-rules/ai-in-the-loop",
     "pie-rules/rule-evolution-methodology",
-    "pie-skills/beginner-friendly-docs",
-    "pie-skills/doc-cross-validator"
+    "pie-skills/beginner-friendly-docs"
   ]
 }

package/assets/public-agent-assets/registry.json

@@ -25,11 +25,11 @@
       "notes": "Public rule for choosing evidence lanes and requiring fresh verification before completion.",
       "promotion": {
         "privateSourcePath": "rules/pie-rules/ai-in-the-loop.md",
-        "privateSourceHash": "sha256:045f9dcd0b1425155a89c863d24ba1509dee4e1e572f23f27fb7fe6f201874ec",
+        "privateSourceHash": "sha256:71f487152123491a2c110ad839a90780b08e04ae3a133d59bbe977bf8b2f64e4",
         "publicHash": "sha256:911702baf6ea1d724e41ea1876d23d80f4de8dfaaadbf1399a40c982c810224c",
         "sanitized": true,
-        "lastReviewed": "2026-06-25",
-        "reviewNotes": "Public copy keeps the evidence-loop method while removing private wording and local-only assumptions."
+        "lastReviewed": "2026-08-17",
+        "reviewNotes": "Reviewed after the private source gained governed front matter and normalized formatting. The public copy intentionally keeps the portable evidence-loop method without the private Grok compatibility lane or project governance metadata."
       }
     },
     {
@@ -87,46 +87,13 @@
       "notes": "Public skill for writing teaching-oriented documentation with stories, analogies, diagrams, and examples.",
       "promotion": {
         "privateSourcePath": "skills/pie-skills/beginner-friendly-docs",
-        "privateSourceHash": "sha256:0098a81bf811f4044a62cbc4ddae30728c219be19df9006df593a2f7b610912e",
-        "publicHash": "sha256:0098a81bf811f4044a62cbc4ddae30728c219be19df9006df593a2f7b610912e",
+        "privateSourceHash": "sha256:1d33061af9376ca30a046c26e2d15f67d2972ea2364e799472fcd31c62ba7104",
+        "publicHash": "sha256:1d33061af9376ca30a046c26e2d15f67d2972ea2364e799472fcd31c62ba7104",
         "sanitized": false,
-        "lastReviewed": "2026-06-25",
+        "lastReviewed": "2026-07-16",
         "reviewNotes": "Public copy is byte-for-byte aligned with the private source at review time."
       },
       "defaultPlacement": "auto"
-    },
-    {
-      "id": "pie-skills/doc-cross-validator",
-      "title": "Doc Cross Validator",
-      "family": "pie-skills",
-      "kind": "skill",
-      "visibility": "public",
-      "sourceKind": "local",
-      "sourcePath": "skills/pie-skills/doc-cross-validator",
-      "hosts": [
-        "codex",
-        "claude-code",
-        "gemini-cli",
-        "antigravity"
-      ],
-      "tags": [
-        "skill",
-        "pie-skills",
-        "documentation",
-        "validation"
-      ],
-      "publishable": true,
-      "origin": "Promoted from maintainer-authored PGS local skills after public review.",
-      "notes": "Public skill for checking documentation claims against source code, config, data structures, and command output.",
-      "promotion": {
-        "privateSourcePath": "skills/pie-skills/doc-cross-validator",
-        "privateSourceHash": "sha256:8aadfc76f9c660ca42ac1dc2c2b25119f8f8328fcf759adb35f9106969411ab4",
-        "publicHash": "sha256:46bd797682876c3973ae6abf757bfc90fcbb3c451ace493eb091ea86e2b2ee0e",
-        "sanitized": true,
-        "lastReviewed": "2026-06-25",
-        "reviewNotes": "Public copy keeps the validation workflow while removing private checkout wording."
-      },
-      "defaultPlacement": "manual"
     }
   ]
 }

package/assets/public-agent-assets/skills/pie-skills/beginner-friendly-docs/SKILL.md

@@ -186,16 +186,14 @@ graph LR
 
 ### 第六步：交叉验证
 
-这一步至关重要，单独列为一个技能（见 `doc-cross-validator` 技能）。
-
-写完后必须对照真实系统验证文档内容，确保没有自己想当然地编造流程。
+这一步至关重要。写完后必须直接对照真实系统的源代码、配置、数据结构和命令输出验证文档内容，确保没有自己想当然地编造流程。
 
 ## 红线（这些是 AI 写教学文档时反复栽的坑，记住"为什么"就不会犯）
 
 **所有层级**：
 - 别用"简单来说"开头——它八成是"我没讲清楚"的遮羞布，真清楚的解释不需要这个前缀
 - 别让一整段纯文字不带任何视觉元素（图/表/引用块）——读者的眼睛需要落脚点，文字墙会劝退人
-- 别写完不做交叉验证——读着顺但流程是错的，比"丑但对"更危险（见 `doc-cross-validator`）
+- 别写完不做交叉验证——读着顺但流程是错的，比"丑但对"更危险
 - 别在没有比喻或例子的情况下抛出复杂概念——抽象概念没有抓手，读者会假装看懂然后悄悄放弃
 
 **初级额外**：

package/assets/starter/AGENTS.template.md

@@ -20,14 +20,28 @@
 - Its selected agents routing file is `PROFILE_ROUTE` under
   `docs/governance/agents-routing/`.
 
+## Policy Discovery
+
+All Markdown under `docs/policy/**/*.md`, including subdirectories and any
+symlinked shared-rule files, belongs to the discoverable policy index. This is
+not a command to load every policy at startup. Read the project-local baseline
+and only the shared rules whose task surface actually matches.
+
+## Skill Availability
+
+An asset manifest or lock records desired state; it does not prove that an
+optional skill is installed, host-discoverable, loaded, or invoked. Use a skill
+only when its SKILL.md actually exists and can be read. Centrally managed
+project links may need the portfolio control plane to materialize them after a
+fresh clone; their absence must not hide or replace the portable policy rules.
+
 ## Documentation Tasks
 
 When the task creates, edits, moves, deletes, or governs documentation, read
 `docs/governance/boundary.md`, `docs/governance/ssot-v1.1.md`,
 `docs/governance/doc-agent-rules.md`, `docs/governance/doc-types.md`, the
-selected agents routing file, and all Markdown under `docs/policy/**/*.md`.
-That policy set includes subdirectories and symlinked shared-rule files; keep
-project AI development policy in `docs/policy/`.
+selected agents routing file, and the policy files that govern the changed
+surface. Keep project AI development policy in `docs/policy/`.
 
 <!-- PGS-ROUTER:END -->
 

package/assets/starter/docs/governance/agents-routing/engineering-runtime-v1.1.md

@@ -6,7 +6,7 @@ status: stable
 canonical: true
 owner: human
 created: 2026-07-13
-last_reviewed: 2026-07-16
+last_reviewed: 2026-08-17
 domain: agents-routing
 tags:
   - agents-routing
@@ -30,13 +30,22 @@ This file decides **how to choose a workflow**, not what the project is currentl
 
 ```mermaid
 flowchart TD
-  A["Task arrives"] --> B["Read project router and current work"]
-  B --> C["Classify with local lane profile"]
-  C --> F["Enter local lane"]
-  F --> H["Use AI-in-the-Loop evidence cycle"]
-  H --> I["Record evidence in doc-gov"]
+  A["Task arrives"] --> B["Read project router and local baseline"]
+  B --> C{"Does the task depend on current priorities or in-flight work?"}
+  C -->|"yes"| D["Read current work"]
+  C -->|"no"| E["Skip current work"]
+  D --> F["Classify with local lane profile"]
+  E --> F
+  F --> G["Enter local lane"]
+  G --> H["Use the relevant verification cycle"]
+  H --> I["Record durable evidence only when its document role requires it"]
 ```
 
+`current-work` is a conditional state surface, not universal startup context.
+Likewise, AI-in-the-Loop applies when implementation or runtime verification
+needs it; a documentation-only or read-only routing task does not load that
+policy by default.
+
 ## Keep It Small
 
 Use this router only to pick depth and workflow. Do not use it as a project roadmap.

package/cli-guide.md

@@ -21,6 +21,9 @@ pro-gov portfolio plan --config /path/to/portfolio.json --target web-app --json
 pro-gov portfolio assets-check --config /path/to/portfolio.json --json
 pro-gov portfolio doctor --config /path/to/portfolio.json --json
 pro-gov portfolio ai-health --config /path/to/portfolio.json --target all --out audits/portfolio-ai-health/2026-07-12
+pro-gov host-lens inspect --config /path/to/portfolio.json --json
+pro-gov host-lens scan --config /path/to/portfolio.json --out audits/host-lens/2026-07-16
+pro-gov host-lens plan --config /path/to/portfolio.json --out audits/host-lens/cleanup-review.json
 pro-gov learn recall --query "release downstream sync" --target .
 pro-gov learn capture --title "Reusable learning" --summary "Reusable lesson..." --target .
 pro-gov lens inspect --target .
@@ -31,9 +34,12 @@ pro-gov init --profile engineering-runtime --dry-run
 pro-gov init --profile engineering-runtime --apply
 pro-gov init --profile doc-only --dry-run
 pro-gov sync --check [--profile engineering-runtime|doc-only]
-pro-gov doctor
+pro-gov package-doctor
 ```
 
+`package-doctor` checks the installed package itself; it is not a target-project
+health claim. The older `doctor` command remains a compatibility alias.
+
 `init --apply` is safe for a fresh target: it preflights every destination and
 refuses the entire operation if any target file already exists. It never merges
 or overwrites an existing router or project-local policy. Existing projects
@@ -73,6 +79,7 @@ pro-gov portfolio plan --config /path/to/control-repo/.pro-gov/portfolio.json --
 pro-gov portfolio assets-check --config /path/to/control-repo/.pro-gov/portfolio.json --json
 pro-gov portfolio doctor --config /path/to/control-repo/.pro-gov/portfolio.json --json
 pro-gov portfolio ai-health --config /path/to/control-repo/.pro-gov/portfolio.json --target all --out audits/portfolio-ai-health/2026-07-12
+pro-gov host-lens scan --config /path/to/portfolio.json --out .scratch/scheduled/host-lens
 pro-gov assets plan --bundle base-governance --target . --out .pro-gov/asset-plan.json
 pro-gov assets plan --bundle project-lens --target /path/to/project --host codex --placement manual --out /tmp/project-lens-plan.json
 pro-gov assets apply --plan .pro-gov/asset-plan.json
@@ -94,7 +101,7 @@ target.
 `portfolio ai-health` performs a read-only structural inventory of the control
 plane, execution engine, registered targets, user/project AI-host configuration,
 exact technology-version policy, the four verification-gate scripts, bounded
-legacy-directory/Playwright-cache evidence, and machine-level specialist checks
+project legacy-directory evidence, and machine-level specialist checks
 explicitly declared in the private manifest (currently DevSpace). It then writes
 a fixed local HTML dashboard plus JSON evidence to the explicit `--out`
 directory. `--target <id>` refreshes one target and merges it into the existing
@@ -102,8 +109,17 @@ same-portfolio board; `--target all` creates complete coverage. Specialist
 checks retain derived health states only; credentials, URLs, and raw process
 environments are never written. The inventory records MCP server names and
 environment filenames, but never reads or emits secret values, environment
-values, MCP commands, arguments, or MCP environment maps. Redundancy evidence is
-read-only and never deletes caches or legacy directories.
+values, MCP commands, arguments, or MCP environment maps. Project redundancy
+evidence is read-only and never deletes legacy directories.
+
+`host-lens` owns host-storage inspection that does not belong in ProjectLens.
+It measures known AI-host roots, package stores, Playwright browsers, explicit
+backup/cache/temp groups, and protected runtime/configuration groups without
+reading file contents. `scan` writes a second fixed SwimmerUIKit dashboard;
+`plan` writes only a review checklist with `executable: false` and
+`approved: false`. Prefer each package manager or host's native maintenance
+tool over generic recursive deletion. See `docs/reference/host-lens.md` in a
+full PGS checkout for the boundary and scheduled-operation contract.
 
 Portfolio manifests are external configuration files owned by a user or
 organization. PGS provides the format and commands; it does not publish a real