kushi-agents 3.4.1

package/plugin/templates/fde/triage-04-readiness-checklist.md

@@ -0,0 +1,82 @@
+# 04 · Mobilization Readiness Checklist — {{project_name}}
+
+> **Bundle file 04 of 07.** Gating view for **mobilization readiness** — can the crew start on day 0 once approval is given?
+>
+> **NOT the same as stage-readiness.** Stage-readiness (separate `fde-report stage-readiness` shape) asks "should we advance the stage?" — mobilization readiness asks "if we got approval today, could the crew actually start?"
+>
+> **Generated by:** Kushi `fde-triage` v{{kushi_version}} on {{generated_at}}.
+
+{{freshness_warning_block}}
+
+## Overall verdict
+
+**{{mobilization_verdict}}**
+
+| Verdict | Meaning |
+|---|---|
+| READY-TO-MOBILIZE | All Hard items ✅ + ≥ 80% of Soft items ✅. Crew can start on day 0. |
+| MOBILIZE-WITH-CAVEATS | All Hard items ✅, but some Soft items ⚠️ or ❓. Document caveats. |
+| BLOCKED | ≥ 1 Hard item ❌ or ❓. Cannot mobilize until resolved. |
+
+This engagement: **Hard items {{hard_pass}}/{{hard_total}}** · **Soft items {{soft_pass}}/{{soft_total}}**.
+
+## Hard items (all required for mobilization)
+
+| # | Item | Status | Evidence | Owner | Source |
+|---|---|---|---|---|---|
+| 1 | Executive sponsor identified and committed | {{h1_status}} | {{h1_evidence}} | {{h1_owner}} | {{h1_source}} |
+| 2 | Customer dev team named (≥ 1 dev, ≥ 1 PM) | {{h2_status}} | {{h2_evidence}} | {{h2_owner}} | {{h2_source}} |
+| 3 | Outcome hypothesis written and agreed | {{h3_status}} | {{h3_evidence}} | {{h3_owner}} | {{h3_source}} |
+| 4 | Quantified success criteria (≥ 2 KPIs with numeric targets + baseline) | {{h4_status}} | {{h4_evidence}} | {{h4_owner}} | {{h4_source}} |
+| 5 | Funding model agreed (MACC / ECIF / billable — line of sight to $) | {{h5_status}} | {{h5_evidence}} | {{h5_owner}} | {{h5_source}} |
+| 6 | CWAA / contracting in place or path to in-place < 30 days | {{h6_status}} | {{h6_evidence}} | {{h6_owner}} | {{h6_source}} |
+| 7 | Catcher team identified (who owns outcome after FDE leaves) | {{h7_status}} | {{h7_evidence}} | {{h7_owner}} | {{h7_source}} |
+| 8 | Azure / MS platform commitment confirmed | {{h8_status}} | {{h8_evidence}} | {{h8_owner}} | {{h8_source}} |
+
+## Soft items (preferred but not blocking)
+
+| # | Item | Status | Evidence | Source |
+|---|---|---|---|---|
+| 1 | Azure landing zone provisioned | {{s1_status}} | {{s1_evidence}} | {{s1_source}} |
+| 2 | Data access path identified (lake / Fabric / pipelines) | {{s2_status}} | {{s2_evidence}} | {{s2_source}} |
+| 3 | Existing PoC or pre-work that informs scope | {{s3_status}} | {{s3_evidence}} | {{s3_source}} |
+| 4 | Industry team alignment confirmed | {{s4_status}} | {{s4_evidence}} | {{s4_source}} |
+| 5 | PG / product team aware (if relevant) | {{s5_status}} | {{s5_evidence}} | {{s5_source}} |
+| 6 | Customer's own ROI estimate documented | {{s6_status}} | {{s6_evidence}} | {{s6_source}} |
+| 7 | Value realization plan with checkpoints (30/60/90 day) | {{s7_status}} | {{s7_evidence}} | {{s7_source}} |
+| 8 | Reusable-pattern flag set (industry-defining / frontier) | {{s8_status}} | {{s8_evidence}} | {{s8_source}} |
+
+## Legend
+
+- ✅ Met — evidence supports
+- ⚠️ Partial — evidence supports some not all
+- ❌ Not met — evidence does NOT support
+- ❓ Unknown — evidence does not address
+
+## Blockers to mobilization
+
+(Sorted by Hard items first.)
+
+{{blockers_block}}
+
+## Required actions to reach READY-TO-MOBILIZE
+
+| # | Action | Owner | By when | Hard/Soft |
+|---|---|---|---|---|
+{{required_actions_rows}}
+
+## Open questions
+
+{{open_questions_block}}
+
+## Validation warnings (this file only)
+
+{{validation_warnings_block}}
+
+## Source coverage
+
+{{source_coverage_block}}
+
+---
+
+> _For "should we advance to the next FDE stage?" use the `stage-readiness` report shape instead — different question._

package/plugin/templates/fde/triage-05-executive-consolidated.md

@@ -0,0 +1,78 @@
+# 05 · Executive Consolidated Report — {{project_name}}
+
+> **Bundle file 05 of 07.** Leadership-friendly combined readout. The single file an exec should read if they only read one.
+> **Generated by:** Kushi `fde-triage` v{{kushi_version}} on {{generated_at}}.
+
+{{freshness_warning_block}}
+
+## Executive Summary
+
+{{executive_summary_paragraph}}
+
+### Key decisions requested from leadership
+
+{{decisions_requested_list}}
+
+### Recommendation
+
+**{{overall_recommendation}}** — {{recommendation_one_liner}}
+
+---
+
+## At a glance
+
+<table>
+<tbody>
+<tr><th>Customer</th><td>{{customer_name}}</td><th>Engagement</th><td>{{engagement_title}}</td></tr>
+<tr><th>FDE stage</th><td>{{current_stage}}</td><th>FDE fit</th><td>{{fit_overall}}</td></tr>
+<tr><th>Mobilization readiness</th><td>{{mobilization_verdict}}</td><th>Top risk</th><td>{{top_risk_short}}</td></tr>
+<tr><th>Customer ROI / yr</th><td>{{customer_roi}}</td><th>Expected MS ACR</th><td>{{acr_expected}}</td></tr>
+<tr><th>MACC</th><td>{{macc_summary}}</td><th>Catcher team</th><td>{{catcher_team}}</td></tr>
+</tbody>
+</table>
+
+## What FDE is being asked to do
+
+{{ask_summary}}
+
+## The outcome hypothesis
+
+> {{outcome_hypothesis}}
+
+## Why FDE here, and not <alternative>
+
+{{why_fde_vs_alternative}}
+
+## Top 3 strengths of this engagement
+
+1. {{strength_1}}
+2. {{strength_2}}
+3. {{strength_3}}
+
+## Top 3 risks / concerns
+
+1. {{risk_1}}
+2. {{risk_2}}
+3. {{risk_3}}
+
+## Asks of leadership
+
+{{leadership_asks}}
+
+## Companion files in this bundle
+
+- `00-fde-analysis.md` — full analysis
+- `01-fde-fitness.md` — 10-criterion scorecard
+- `02-risk-analysis.md` — full risk view
+- `03-6Q.md` — engagement framing
+- `04-readiness-checklist.md` — mobilization readiness
+- `06-global-opportunity-and-reuse.md` — repeatability lens
+- `07-validation-warnings-checklist.md` — open warnings tracker
+
+## Validation warnings (this file only — full list in `07-`)
+
+{{validation_warnings_block}}
+
+## Source coverage
+
+{{source_coverage_block}}

package/plugin/templates/fde/triage-06-global-opportunity.md

@@ -0,0 +1,70 @@
+# 06 · Global Opportunity & Reuse — {{project_name}}
+
+> **Bundle file 06 of 07.** Repeatability, reuse, and commercialization view. Grounded in project Evidence plus public context where the broader-opportunity lens applies.
+> **Generated by:** Kushi `fde-triage` v{{kushi_version}} on {{generated_at}}.
+
+{{freshness_warning_block}}
+
+## Is this engagement pattern reusable?
+
+**{{reusability_verdict}}** — {{reusability_one_liner}}
+
+| Verdict | Meaning |
+|---|---|
+| HIGH-REUSE | Pattern repeats across ≥ 5 known customers in same industry / use case. Strong PG accelerator candidate. |
+| MEDIUM-REUSE | Pattern likely repeats across 2–4 customers. Worth a reusable accelerator after this engagement. |
+| LOW-REUSE | Custom to this one customer. Unlikely to inform a broader pattern. |
+| UNKNOWN | Insufficient evidence to assess. |
+
+## Industry / segment lens
+
+| Field | Value | Source |
+|---|---|---|
+| Industry | {{industry}} | {{industry_source}} |
+| Sub-segment | {{sub_segment}} | {{sub_segment_source}} |
+| Customer size (revenue / employees) | {{customer_size}} | {{customer_size_source}} |
+| Geography | {{geography}} | {{geography_source}} |
+
+## Pattern description
+
+{{pattern_paragraph}}
+
+## Similar customers / similar problems we know of
+
+| Customer | How similar | Status (FDE / partner / un-engaged) | Source |
+|---|---|---|---|
+{{similar_customers_rows}}
+
+(If from public context, cite the public source URL.)
+
+## Reusable artifacts this engagement could produce
+
+| Artifact | Reuse target | Description |
+|---|---|---|
+{{reusable_artifacts_rows}}
+
+## Commercial opportunity at scale
+
+If this pattern repeats across N similar customers:
+
+- Estimated TAM (Microsoft ACR uplift): {{tam_acr}}
+- Estimated number of similar customers: {{n_similar}}
+- Source for sizing: {{tam_source}}
+
+## PG / accelerator handoff candidates
+
+| PG team / accelerator | Why this engagement is relevant | Source |
+|---|---|---|
+{{pg_handoff_rows}}
+
+## Risks to reuse
+
+{{reuse_risks_block}}
+
+## Validation warnings (this file only)
+
+{{validation_warnings_block}}
+
+## Source coverage
+
+{{source_coverage_block}}

package/plugin/templates/fde/triage-07-validation-warnings.md

@@ -0,0 +1,60 @@
+# 07 · Validation Warnings Checklist — {{project_name}}
+
+> **Bundle file 07 of 07.** Central tracker of every validation warning across the triage bundle.
+> Warnings are mirrored here verbatim from companion files; updates to status (open / resolved / not-applicable) live here.
+>
+> **Generated by:** Kushi `fde-triage` v{{kushi_version}} on {{generated_at}}.
+
+## Summary
+
+| Status | Count |
+|---|---|
+| 🔴 Open | {{count_open}} |
+| 🟢 Resolved | {{count_resolved}} |
+| ⚪ Not applicable | {{count_na}} |
+
+**Total warnings across bundle:** {{count_total}}
+
+## Open warnings (blockers / action-needed)
+
+| # | Bundle file | Rule | Description | Status | Owner | Action |
+|---|---|---|---|---|---|---|
+{{open_warnings_rows}}
+
+## Resolved warnings
+
+| # | Bundle file | Rule | Description | Resolved on | Resolution note |
+|---|---|---|---|---|---|
+{{resolved_warnings_rows}}
+
+## Not-applicable warnings
+
+| # | Bundle file | Rule | Description | Justification for NA |
+|---|---|---|---|---|
+{{na_warnings_rows}}
+
+## Rule reference
+
+[Source: reference-packs/fde/report-doctrine.md · packaged]
+
+| Rule | Title |
+|---|---|
+| Rule 1.1 | Source resolution gate skipped for a required path |
+| Rule 1.2 | Source attempted-failed for a path material to a claim |
+| Rule 2.1 | Recency precedence violated (newer source ignored, older quoted) |
+| Rule 3.1 | CRM field value reported as confirmed fact without customer cross-check |
+| Rule 3.2 | CRM disagrees with latest customer-facing source |
+| Rule 4.1 | Required role (ATU rep, Industry rep) missing |
+| Rule 4.2 | Required commercial fact missing |
+| Rule 4.3 | Required outcome metric (KPI, baseline, target) missing |
+| Rule 5.1 | Stage claim not supported by stage gate / exit criteria |
+| Rule 5.2 | FDE fit claim not grounded against the 10-row fitness checklist |
+| Rule 6.1 | Customer engagement legal posture (CWAA / EMA / contract) absent |
+| Rule 8.1 | Paragraph lacks inline `[source: …]` citation |
+| Rule 9.1 | Other unresolved gap (free-text) |
+
+## How to update this file
+
+This file is **regenerated** every time `@Kushi fde-triage <project>` runs. To preserve status changes (open → resolved, NA justifications), edit and re-save **before** re-running triage — `fde-triage` merges existing status when found.
+
+Or update individual warnings inline in the source bundle file (00 — 06) and re-run.

package/plugin/templates/init/ado-config.template.yml

@@ -0,0 +1,21 @@
+# Azure DevOps connection — per-user filled. Lives at:
+#   <engagement-root>/.project-evidence/ado/config.yml
+# NOT in repo. NOT in `.kushi/`.
+
+tenantId: "72f988bf-86f1-41af-91ab-2d7cd011db47"   # Microsoft tenant
+organization: "https://dev.azure.com/<your-org>"
+defaultProject: "<your-default-ado-project>"
+apiVersion: "7.1"
+
+# ADO scope GUID — fixed across tenants
+azDevOpsResourceId: "499b84ac-1321-427f-aa17-267ca6975798"
+
+# Saved queries / area paths can be overridden per project in
+#   <engagement-root>/<project>/integrations.yml under `ado:`
+defaults:
+  queryId: ""
+  areaPath: ""
+  iterationPath: ""
+
+auth:
+  strategy: "az-cli"

package/plugin/templates/init/crm-config.template.yml

@@ -0,0 +1,16 @@
+# CRM (Dataverse) connection — per-user filled. Lives at:
+#   <engagement-root>/.project-evidence/crm/config.yml
+# NOT in repo. NOT in `.kushi/`.
+
+tenantId: "72f988bf-86f1-41af-91ab-2d7cd011db47"   # Microsoft tenant
+environmentUrl: "https://<your-org>.crm.dynamics.com"
+apiVersion: "9.2"
+
+# Default entity to query. Override per project in
+#   <engagement-root>/<project>/integrations.yml
+defaultEntitySet: "msdyn_engagements"
+
+# Authentication strategy. Currently only "az-cli" supported.
+# kushi will run: az account get-access-token --resource <environmentUrl> --tenant <tenantId>
+auth:
+  strategy: "az-cli"

package/plugin/templates/init/kushi-projects.template.json

@@ -0,0 +1,14 @@
+{
+  "$schema": "https://kushi.dev/schemas/projects.v2.json",
+  "_comment": "Project list consumed by the Clawpilot automation 'Kushi weekly refresh'. Copy this template to ~/.copilot/kushi-projects.json and edit `projects[]`. Per project: set `force_refresh: true` to force a full re-pull on the NEXT run (ignores the run-log watermark) — the automation flips it back to false after a successful run. Set `mode: \"aggregate\"` for Evidence-only (pull + consolidate, no State/ rebuild) or `\"refresh\"` (default; pull + consolidate + build-state). The `aggregate` mode requires Kushi installed with the `core` profile or higher. To trigger on demand, set the flag(s) and tell Clawpilot 'kushi run again now' — it will call m_run_automation_now on the weekly automation.",
+  "version": 2,
+  "defaults": {
+    "force_refresh": false,
+    "window_days": 7,
+    "mode": "refresh"
+  },
+  "projects": [
+    { "name": "<ProjectA>", "mode": "refresh",   "force_refresh": false },
+    { "name": "<ProjectB>", "mode": "aggregate", "force_refresh": false }
+  ]
+}

package/plugin/templates/init/m365-auth.template.json

@@ -0,0 +1,67 @@
+{
+  "_meta": {
+    "owner": "<alias>@microsoft.com",
+    "purpose": "STABLE per-user M365 defaults for Microsoft consultants. Tenant is Microsoft corp by default. Edit by hand only when defaults change. NOT auto-overwritten by skill runs.",
+    "consumers": ["project-evidence skill", "any future M365-context skill"],
+    "audience": "Microsoft internal — Industry Solutions Engineering and adjacent teams",
+    "schema_version": "1.0",
+    "last_reviewed": "<YYYY-MM-DD>"
+  },
+  "m365Auth": {
+    "defaultTenantId": "72f988bf-86f1-41af-91ab-2d7cd011db47",
+    "resources": {
+      "graph": "https://graph.microsoft.com",
+      "sharePoint": "https://microsoft-my.sharepoint.com",
+      "dataverse": "https://iscrm.crm.dynamics.com"
+    },
+    "oneNote": {
+      "defaultNotebookName": "",
+      "defaultNotebookId": "",
+      "defaultSectionResolverUrl": "",
+      "defaultNotebookRootLink": "",
+      "defaultLinkOwner": ""
+    },
+    "emailContext": {
+      "enabled": true,
+      "dateFloor": "",
+      "folders": [],
+      "includeSubfolders": true,
+      "sourceCoverageLabel": "",
+      "matchingPolicy": {
+        "mode": "hybrid",
+        "rankingOrder": ["exact", "prefix", "contains"],
+        "minConfidenceForFolderScopedSearch": "high",
+        "fallbackToFullRootScanWhenAmbiguous": true,
+        "alwaysFuzzy": true
+      }
+    },
+    "teamsChatContext": {
+      "enabled": true,
+      "dateFloor": "",
+      "scope": { "includeChats": true, "includeChannels": true },
+      "matchingPolicy": {
+        "mode": "thread-first",
+        "rankingOrder": ["exact", "prefix", "contains"],
+        "fallbackToBroaderSearchWhenAmbiguous": true,
+        "alwaysFuzzy": true
+      }
+    },
+    "calendarContext": {
+      "enabled": true,
+      "matchingPolicy": {
+        "mode": "subject-and-body-keywords",
+        "rankingOrder": ["exact", "prefix", "contains"],
+        "alwaysFuzzy": true
+      }
+    },
+    "sharePointContext": {
+      "enabled": true,
+      "localProjectsRoot": "",
+      "matchingPolicy": {
+        "mode": "fuzzy-folder-name",
+        "rankingOrder": ["exact", "prefix", "contains"],
+        "alwaysFuzzy": true
+      }
+    }
+  }
+}

package/plugin/templates/init/m365-mutable.template.json

@@ -0,0 +1,19 @@
+{
+  "_meta": {
+    "owner": "<alias>@microsoft.com",
+    "purpose": "MUTABLE per-project operational mappings (Microsoft internal). Auto-updated by the project-evidence skill at the end of every weekly run when high-confidence hints are discovered. Safe to hand-edit. New project entries are added under m365Mutable.knownSections during B2 bootstrap and populated incrementally as the skill discovers folders, sections, chats, etc.",
+    "audience": "Microsoft internal",
+    "matching": "case-insensitive on project key (exact > prefix > contains); fuzzy match always on",
+    "schema_version": "1.0",
+    "last_updated": "<YYYY-MM-DD>"
+  },
+  "m365Mutable": {
+    "metadata": {
+      "description": "Mutable M365 operational mappings used to speed and stabilize context retrieval.",
+      "defaultNotebookRef": "../m365/m365-auth.json -> m365Auth.oneNote.defaultNotebookId",
+      "matching": "case-insensitive by project key (exact/prefix/contains)",
+      "source": "Derived from project folder names, mailbox/Teams/OneNote discovery during bootstrap/refresh/retry runs."
+    },
+    "knownSections": {}
+  }
+}

package/plugin/templates/init/project-contributors.template.yml

@@ -0,0 +1,27 @@
+# Contributors to the {{PROJECT_NAME}} project Evidence folder.
+# Each contributor owns a subfolder named after their alias and is responsible for
+# their own per-source extracts (Email, Teams-Chats, OneNote, Meetings).
+# SharePoint extracts can be produced by any contributor (folder is shared).
+
+project: {{PROJECT_NAME}}
+project_folder: {{PROJECT_FOLDER_PATH}}
+week_definition: Monday-Sunday
+
+contributors:
+  - alias: {{alias}}
+    name: {{Full Name}}
+    email: {{email@domain.com}}
+    role: {{role}}
+    sources_owned: [Email, Teams-Chats, OneNote, SharePoint, Meetings]
+    last_run: null
+    notes: ""
+
+consolidation:
+  output_folder: _Consolidated
+  cadence: weekly
+  owner: {{alias}}   # the person who runs the merge step
+  rules:
+    - Union decisions, action items, open questions; de-duplicate by semantic match.
+    - Tag every fact with [source: <alias>/<source-folder>] for provenance.
+    - Surface conflicts in a "Conflicts / Discrepancies" section rather than silently picking one.
+    - Do NOT include other contributors' raw mailbox/chat content the consolidator can't see; rely on their summaries.

package/plugin/templates/init/project-evidence.template.yml

@@ -0,0 +1,32 @@
+# Personal config for Kushi.
+# Copy to:  ~/.copilot/project-evidence.yml   (Linux/macOS)
+#           %USERPROFILE%\.copilot\project-evidence.yml   (Windows)
+#
+# Each team member maintains their own copy. Nothing in here is a secret,
+# but it IS personal — never commit a filled copy to a shared repo.
+
+# Short id used as your evidence subfolder name (e.g. "alex", "jordan").
+alias: <your-alias>
+
+display_name: <Your Full Name>
+email: your.email@example.com
+
+# Where your engagement-root lives — i.e. the parent folder that contains
+# one subfolder per project (typically synced from a team SharePoint library).
+# Examples:
+#   Windows : 'C:\Users\<you>\OneDrive - <Tenant>\<Team>\Engagement Assets'
+#   macOS   : '/Users/<you>/OneDrive - <Tenant>/<Team>/Engagement Assets'
+projects_root: '<engagement-root>'
+
+# Default OneNote notebook to scan for project pages.
+default_onenote_notebook: <Your Notebook Name>
+
+# Projects you actively work on. Kushi scans only these unless you ask otherwise.
+# Match is fuzzy against subfolder names under projects_root.
+active_projects:
+  - <ProjectA>
+  # - <ProjectB>
+  # - <add more here>
+
+# (Optional) where the global integrations live, if not the default.
+# integrations_root: '<USER_HOME>/.copilot/integrations'

package/plugin/templates/init/project-integrations.template.yml

@@ -0,0 +1,34 @@
+# Per-project integration filter.  Lives at: <project>\integrations.yml
+# Tells the skill how to find THIS project's records in the (global) CRM and ADO.
+# Connection / auth / field mappings are NOT here — those are global.
+
+project: '{{PROJECT_NAME}}'
+
+crm:
+  # Provide ONE of these:
+  record_id: ''                          # e.g. 00000000-0000-0000-0000-000000000000  (preferred)
+  request_id: ''                         # e.g. REQ-12345
+  title_contains: '{{PROJECT_NAME}}'     # last-resort fuzzy
+  # Optional: override default app context
+  # app_id: 833534b8-5db0-4036-a029-4e4e4d7d4315
+
+ado:
+  # Provide ONE primary locator (more specific = better):
+  engagement_id: 0                       # ADO Engagement work item ID  (preferred)
+  area_path: ''                          # e.g. 'IS Engagements\\<Project>'
+  tag: ''                                # e.g. 'noridian'
+  title_filter: '{{PROJECT_NAME}}'       # used by items_changed_between WIQL
+  # Optional: ADO query you've saved
+  # query_id: ''
+
+sharepoint:
+  # Optional online SharePoint site for this project (in addition to / instead of a local synced folder).
+  # If set, the SharePoint source pulls file metadata + activity via Graph (drives/items/lastModifiedDateTime)
+  # instead of (or in addition to) walking the local sync folder.
+  # Add via bootstrap, OR later by saying "add sharepoint <URL> to <project>".
+  sites:
+    # - url: 'https://contoso.sharepoint.com/sites/MyProject'    # full site URL
+    #   library: 'Documents'                                      # document library name (default: Documents)
+    #   folder: ''                                                # optional sub-folder path inside the library
+    #   include_subfolders: true
+    #   recurse_depth: 5

package/plugin/templates/init/project-user-settings.template.yml

@@ -0,0 +1,71 @@
+# Per-(project × user) settings.
+# Lives at: <project>\Evidence\<alias>\.settings.yml
+# Created and maintained by the bootstrap step (BOOTSTRAP.md).
+# Anything left blank below will trigger a prompt next time the skill runs for this (project × user) pair.
+# Nothing in here is secret — it's per-user query parameters and run state, not credentials.
+
+project: '{{PROJECT_NAME}}'
+alias: '{{ALIAS}}'
+display_name: '{{FULL_NAME}}'
+email: '{{EMAIL}}'
+
+# ── Per-user query scope (your data only) ─────────────────────────────────────
+mailbox:
+  folder: 'inbox'                         # well-known folder name or custom folder name
+  search_keywords: []                     # extra search terms to OR with project name (e.g. ['MPXe','policy comments'])
+
+teams_chats:
+  # Specific Teams chat IDs known to be project-relevant for THIS user.
+  # Leave empty to fall back to keyword search across recent chats.
+  chat_ids: []                            # e.g. ['19:meeting_abc...@thread.v2']
+  search_keywords: []
+
+onenote:
+  notebook: '{{ONENOTE_NOTEBOOK}}'        # default from ~\.copilot\project-evidence.yml
+  section: '{{ONENOTE_SECTION}}'          # e.g. 'Noridian.one'
+  page_filter: ''                         # optional substring filter
+
+calendar:
+  organizer_filter: ''                    # optional — restrict to meetings organized by these emails
+  subject_keywords: []                    # extra subject search terms beyond project name
+
+sharepoint:
+  # Per-user filter on top of the project-shared sites in <project>\integrations.yml.
+  # Leave empty to include everything from the project's configured sites.
+  author_filter: []                       # restrict to files authored/last-modified-by these emails (e.g. ['{{EMAIL}}'])
+  path_includes: []                       # restrict to paths containing any of these substrings
+  path_excludes: ['_Archive', 'Old']      # skip paths containing these substrings
+  snapshot:
+    # Snapshot pass body-extraction depth. Tree.md ALWAYS captures full hierarchy.
+    # These knobs only control which files get a /snapshot/files/<path>.md body summary.
+    recent_n: 25                          # capture body summaries for the top-N most-recently-modified files; 0 disables
+    always_include: []                    # additional relative paths or substrings to always extract (e.g. ['Architecture/', 'SOW.docx'])
+
+# ── Q&A behavior (ask-project skill) ──────────────────────────────────────────
+chat:
+  freshness_warn_days: 14                   # warn before answering if relevant source last-pulled > N days ago; never auto-refreshes
+  default_state_file: 'State/00_overview.md' # fallback file when question doesn't map to a specific State topic
+  redact_attendee_emails_in_answers: true   # strip raw emails from chat answers (names OK)
+  max_files_per_answer: 8                   # hard cap on files loaded per answer to keep responses focused
+  offer_refresh_on_stale: true              # if stale, offer @Kushi refresh — but never auto-execute it
+
+# ── Per-user privacy / outbound ───────────────────────────────────────────────
+outbound:
+  allow_proactive_send: false             # HARD DEFAULT: never send anything without explicit per-turn approval
+  notify_on_finding: false                # if true, suggest (but do not send) outbound notifications when conflicts surface
+  preferred_review_doc: 'State/09_open-questions.md'
+
+# ── Per-user run state ────────────────────────────────────────────────────────
+last_runs:                                # populated by the skill after each run
+  email: null
+  teams_chats: null
+  onenote: null
+  sharepoint: null
+  meetings: null
+  crm: null
+  ado: null
+
+# ── Per-user role on this project (optional) ──────────────────────────────────
+role: ''                                   # e.g. 'engagement lead', 'ISE engineer', 'observer'
+sources_owned: [Email, Teams-Chats, OneNote, SharePoint, Meetings]
+notes: ''

package/plugin/templates/paste-prompt.md

@@ -0,0 +1,35 @@
+<!--
+Paste-prompt template — emitted by pull-* skills when WorkIQ + retry both fail to produce
+a thoroughness-compliant evidence file. The skill MUST customize the bracketed fields and
+present this block verbatim to the user, then wait for paste.
+-->
+
+⚠️ **Thoroughness check failed for `<evidence-file-path>`**
+
+WorkIQ returned thin content for this source even after a deeper-query retry. To meet the evidence thoroughness bar (full bodies, not just metadata), I need you to paste the source content directly.
+
+**Failing checks:**
+- `<check-id-1>` — `<one-line reason>`
+- `<check-id-2>` — `<one-line reason>`
+
+**What to paste**
+
+For `<source>` (`<entity-name>`), open the source in `<Outlook / Teams / OneNote / SharePoint / CRM / ADO>` and paste:
+
+- `<source-specific item 1, e.g. "Full email body of the 5 threads listed above">`
+- `<source-specific item 2, e.g. "Reply chain for each thread, oldest → newest">`
+- `<source-specific item 3, e.g. "Any attachments referenced">`
+
+**Where to paste**
+
+Paste between the markers below. I'll incorporate it into `<evidence-file-path>` and re-run the detector.
+
+```
+<<< PASTE START
+... your paste here ...
+PASTE END >>>
+```
+
+**Skip this entity?**
+
+If this content genuinely isn't retrievable (deleted, permission-blocked, out of scope), reply `skip <entity-name>` and I'll mark the entry `[tree-only]` with a coverage note.