kushi-agents 4.8.2 → 4.9.0

package/plugin/instructions/weekly-csc.instructions.md

@@ -0,0 +1,335 @@
+---
+applyTo: "**"
+description: "Weekly CSC layout (kushi v4.9.0+) — every source produces ONE weekly CSC file per ISO-week containing one block per entity touched that week, plus a per-source _index/entities.yml for current-state lookup. Replaces the snapshot/ + stream/ folder split."
+---
+
+# Weekly CSC Layout — HARD RULE (kushi v4.9.0+)
+
+## Why this replaces snapshot/ + stream/
+
+The old `snapshot-vs-stream.instructions.md` model split every source into two shapes:
+- `snapshot/` — current state of entities (replace on refresh)
+- `stream/` — timestamped events (append, weekly buckets)
+
+In practice this caused two problems:
+1. **Empty snapshot folders.** Sources without retrievable bodies (OneNote payload-too-large,
+   m365_get_email failures, no SharePoint site discovery) ended up with no `snapshot/`
+   content. Users perceived empty folders as "kushi didn't pull anything".
+2. **Awkward queries.** Answering "what happened on HCA last week?" required reading
+   `stream/<week>.md` per source AND walking `snapshot/` to know current state of entities
+   touched that week. Six sources × two folder reads per query.
+
+The CSC model (`comprehensive-structured-capture.instructions.md`) makes every per-entity
+block self-contained — it carries current state at end-of-week + everything that happened to
+the entity that week. So one folder is enough: organize by week.
+
+## The layout
+
+```
+<engagement-root>/<project>/Evidence/<alias>/
+   email/
+      weekly/2026-05-25_email-csc.md
+      weekly/2026-05-18_email-csc.md
+      _index/entities.yml
+   teams/
+      weekly/2026-05-25_teams-csc.md
+      _index/entities.yml
+   meetings/
+      weekly/2026-05-25_meetings-csc.md
+      verbatim/2026-05-21-1400_hca-weekly-sync/
+         captured-at.txt
+         chat-messages.json
+         chat-messages.md
+         transcript.txt          (or transcript.vtt, or transcript-source.md)
+         recording-url.txt
+         recap-card.md
+         attachments/
+         coverage.md
+      _index/entities.yml
+   onenote/
+      weekly/2026-05-25_onenote-csc.md
+      _index/entities.yml
+   sharepoint/
+      weekly/2026-05-25_sharepoint-csc.md
+      _index/entities.yml
+   crm/
+      weekly/2026-05-25_crm-csc.md
+      _index/entities.yml
+   ado/
+      weekly/2026-05-25_ado-csc.md
+      _index/entities.yml
+   run-log.yml
+   refresh-reports/
+      2026-05-26-1830_bootstrap.md
+      ...
+```
+
+Notes:
+- **One weekly file per source per ISO-week.** Filename = `YYYY-MM-DD_<source>-csc.md` where
+  the date is the **Monday** of the ISO-week the entries cover (Mon–Sun).
+- **Empty weeks produce no file.** If a source had no activity in a week, no file is written.
+- **`meetings/verbatim/<dir>/` is the only sub-folder besides weekly/ + _index/.** It is the
+  raw audit folder per `meetings-verbatim-required.instructions.md` (meetings are the only
+  source whose raw evidence expires; tenant retention purges Teams recordings, often within
+  60 days).
+- **No `snapshot/` and no `stream/` folders are written by v4.9.0+ skills.** Legacy folders
+  from prior kushi versions are left in place (no migration); readers fall back to them when
+  `weekly/` is empty (see Reader contract below).
+
+## Weekly file structure
+
+```markdown
+# <Source> CSC — Week of 2026-05-25 (Mon–Sun)
+
+- **Project**: <project>
+- **Alias**: <alias>
+- **Source**: <email | teams | meetings | onenote | sharepoint | crm | ado>
+- **Week**: 2026-05-25 → 2026-05-31 (ISO)
+- **Generated**: <ISO timestamp>
+- **Kushi version**: <version>
+- **WorkIQ calls**: <count> (request-ids listed under each entity)
+
+## Source Basis (week-level)
+
+- <one paragraph describing what WorkIQ probes / host probes ran this week,
+   what scope was applied (boundaries), what failed, what was deferred>
+- See per-entity Source basis lines for granular provenance.
+
+## Entities
+
+<one CSC block per entity touched this week, in chronological order of last_touched>
+
+## <entity-1-name> {#<entity-1-anchor>}
+...CSC block per `comprehensive-structured-capture.instructions.md`...
+
+## <entity-2-name> {#<entity-2-anchor>}
+...
+
+## Week-level Coverage Notes
+
+- <anything that affects the whole week, not specific to one entity>
+- <deferred-retry markers written this week (cite the marker paths)>
+
+## Validation
+
+- [ ] every entity block has the canonical section order
+- [ ] no prose paragraphs anywhere
+- [ ] no truncation markers ("and N more", etc.)
+- [ ] every fact-bearing bullet has a citation
+- [ ] `_index/entities.yml` updated for every entity in this file
+```
+
+## `_index/entities.yml` shape
+
+One per source per contributor. Maps entity-id → location of latest CSC mention. Used by
+`build-state`, `ask-project`, and `consolidate-evidence` for current-state lookup without
+walking every weekly file.
+
+```yaml
+# Auto-generated. Updated by pull-<source> on every write.
+# Per-contributor (under Evidence/<alias>/<source>/_index/). Multi-user safe.
+schema_version: 1
+source: onenote                            # one of: email|teams|meetings|onenote|sharepoint|crm|ado
+generated_at: 2026-05-26T18:00:00Z
+kushi_version: 4.9.0
+entities:
+  # Stable entity-id chosen per source (see "Entity identity" below).
+  - id: 'onenote://page/wdpartid=abc123'
+    display_name: 'HCA Architecture Overview'
+    entity_anchor: 'hca-architecture-overview'
+    latest_csc_file: 'weekly/2026-05-25_onenote-csc.md'
+    latest_csc_block_offset: 412               # line number of the entity heading
+    last_touched: 2026-05-23T14:30:00Z
+    first_seen: 2026-05-06T09:12:00Z
+    weeks_touched:
+      - 2026-05-25
+      - 2026-05-18
+      - 2026-05-04
+    status: captured                           # captured | body-not-exposed | unavailable | deferred
+  - id: 'meetings://joinurl=https%3A%2F%2Fteams.microsoft.com%2F...'
+    display_name: 'HCA Weekly Sync'
+    entity_anchor: 'hca-weekly-sync-2026-05-21-1400'
+    latest_csc_file: 'weekly/2026-05-25_meetings-csc.md'
+    latest_csc_block_offset: 95
+    last_touched: 2026-05-21T14:00:00Z
+    first_seen: 2026-04-23T14:00:00Z
+    weeks_touched: [2026-05-18, 2026-05-04, 2026-04-27, 2026-04-20]
+    verbatim_folder: 'meetings/verbatim/2026-05-21-1400_hca-weekly-sync'
+    status: captured
+```
+
+### Entity identity per source (canonical id format)
+
+| Source | Entity | id format |
+|---|---|---|
+| email | thread | `email://conversation_id=<conversationId>` |
+| teams | chat thread | `teams://chat_id=<chatId>` |
+| teams | channel thread | `teams://channel=<teamId>/<channelId>/<rootMessageId>` |
+| meetings | series instance | `meetings://joinurl=<urlencoded>` + `&start=<iso>` |
+| onenote | page | `onenote://page/wdpartid=<id>` (preferred) or `onenote://page/webPageId=<id>` |
+| sharepoint | file | `sharepoint://siteurl=<urlencoded>/path=<urlencoded>` or `sharepoint://driveitemid=<id>` |
+| crm | record | `crm://entity=<logicalName>/id=<guid>` |
+| ado | work item | `ado://org=<org>/project=<project>/workitemid=<int>` |
+
+These ids are stable across runs and contributors. Multi-user safety: every contributor writes
+their own `_index/entities.yml` under their own `<alias>/` folder; no shared file.
+
+## Reader contract (build-state, ask-project, consolidate-evidence)
+
+Readers MUST follow this fallback chain to support graceful migration from pre-v4.9.0
+evidence:
+
+1. **Primary**: read `Evidence/<alias>/<source>/_index/entities.yml`. Pull `latest_csc_file`
+   for each entity needed. Read the CSC block by anchor.
+2. **Secondary (no _index/)**: walk `Evidence/<alias>/<source>/weekly/*.md` newest-first;
+   parse entity headings; build an in-memory index for this query.
+3. **Legacy fallback (pre-v4.9.0)**: if `weekly/` is empty BUT `snapshot/` or `stream/` exist,
+   read those instead. Cite as legacy: `[source: <alias>/<source>/snapshot/<file>.md (legacy
+   pre-v4.9.0 layout)]`. Do not refuse to read legacy.
+4. Mark the answer with `Source-layout: weekly-csc` / `weekly-csc-plus-legacy` /
+   `legacy-only` in the answer footer so the user knows if a refresh would update the layout.
+
+Readers MUST NOT delete or modify legacy folders. Migration is opt-in via a future migrate
+skill (out of scope for v4.9.0).
+
+## Writer contract (every pull-* skill)
+
+When writing weekly CSC:
+
+1. Compute the ISO-week Monday for each entity touched (entity may span weeks — assign each
+   touch to its own week's file).
+2. For each week-bucket:
+   - Open `Evidence/<alias>/<source>/weekly/<monday>_<source>-csc.md`. Create if absent.
+   - For each entity in this week-bucket:
+     - If entity heading `## <name> {#<anchor>}` exists → **merge** the CSC block:
+       - Replace the existing block in place (current week's CSC always reflects end-of-week
+         state).
+     - Else → **append** a new CSC block at end of `## Entities` section.
+   - Update week-level Source Basis + Coverage Notes.
+3. Update `Evidence/<alias>/<source>/_index/entities.yml`:
+   - For each entity written: upsert the row; bump `last_touched`; add this week's Monday to
+     `weeks_touched`; set `latest_csc_file` + `latest_csc_block_offset`.
+4. Append a `runs:` entry to `Evidence/run-log.yml` per `auth-and-retry.instructions.md`.
+
+### Idempotency rules
+
+- Same entity, same week, two runs → second run **replaces** the block in place. (The CSC
+  block always reflects end-of-week state; intra-week reruns refresh the same block.)
+- Same entity, different weeks → each week's file gets its own block.
+- Deletion of a previously-captured entity → leave prior weeks' blocks alone (history is
+  preserved by the weekly partition). Mark `status: deleted-from-source` in `_index/` and
+  do NOT write a block in the current week.
+
+### Multi-user safety
+
+`Evidence/<alias>/...` is per-contributor. The `<alias>/_index/entities.yml` is
+**not shared** across contributors. Cross-contributor consolidation is the job of
+`consolidate-evidence` (which reads each contributor's `_index/` independently). No merge
+conflicts possible at the writer level.
+
+## Examples
+
+### Email weekly file (excerpt)
+
+```markdown
+# Email CSC — Week of 2026-05-25 (Mon–Sun)
+
+- Project: HCA
+- Alias: ushak
+- Source: email
+- Week: 2026-05-25 → 2026-05-31
+- Generated: 2026-05-26T18:45:00Z
+- Kushi version: 4.9.0
+- WorkIQ calls: 1 enumeration + 6 thread CSC
+
+## Source Basis
+
+- WorkIQ enumeration call request-id abc123 returned 6 in-scope threads in folder
+   "Customers/HCA" between 2026-05-25 and 2026-05-31.
+- 6 thread-level WorkIQ CSC calls returned full bulleted capture; no deferrals.
+
+## Entities
+
+## HCA Compute Architecture Discussion {#hca-compute-arch-thread}
+
+- Source basis: WorkIQ request-id def456, fidelity csc-full
+- Coverage window: 2026-05-25 → 2026-05-30
+- Last touched: 2026-05-30T09:12:00Z
+
+### Participants / Present
+- John Doe <john@hca.com> — sender (initiated thread)
+- Sample Name <you@example.com> — recipient + responder
+- ...
+
+### Topics Discussed
+- Migration strategy for legacy compute workloads
+  - Current state: 200 VMs on-prem, Hyper-V
+  - Discussed: lift-and-shift vs refactor-to-AKS trade-offs
+- ...
+
+### Decisions
+- Decided: pilot lift-and-shift for non-critical 30-VM workload first — by John Doe + Usha
+  K — rationale: lower risk for first migration; refactor decision deferred until pilot
+  results — rejected: full refactor in first wave [source: WorkIQ req def456 · 2026-05-28]
+
+### Dates & Numbers Shared
+- Pilot start date: 2026-06-15 — context: 30-VM lift-and-shift kickoff — shared by John Doe
+   [source: WorkIQ req def456 · 2026-05-28]
+- Budget cap: $250K — context: pilot phase only — shared by Mike Smith [source: WorkIQ req
+   def456 · 2026-05-29]
+
+### Action Items
+- [ ] John Doe — confirm pilot VM list with infrastructure team — due: 2026-06-05 [source:
+   WorkIQ req def456 · 2026-05-29]
+- [ ] Usha Krishnan — share migration playbook deck — due: 2026-05-31 [source: WorkIQ req
+   def456 · 2026-05-30]
+
+...
+```
+
+### `_index/entities.yml` after that write
+
+```yaml
+schema_version: 1
+source: email
+generated_at: 2026-05-26T18:45:00Z
+kushi_version: 4.9.0
+entities:
+  - id: 'email://conversation_id=AAQkAGI...'
+    display_name: 'HCA Compute Architecture Discussion'
+    entity_anchor: 'hca-compute-arch-thread'
+    latest_csc_file: 'weekly/2026-05-25_email-csc.md'
+    latest_csc_block_offset: 28
+    last_touched: 2026-05-30T09:12:00Z
+    first_seen: 2026-05-25T11:00:00Z
+    weeks_touched: [2026-05-25]
+    status: captured
+  - id: 'email://conversation_id=AAQkAGI...other...'
+    ...
+```
+
+## Self-check enforcement
+
+`plugin/skills/self-check/run.ps1` deep checks added in v4.9.0:
+- **D16**: `_index/entities.yml` exists when `weekly/` is non-empty.
+- **D17**: every entity heading in a weekly file has a matching `_index/` row, and vice versa.
+- **D18**: entity ids in `_index/` follow the canonical format for the source.
+- **D19**: warn (not fail) on `snapshot/` or `stream/` folders inside `Evidence/<alias>/<source>/`
+  written by kushi ≥ v4.9.0 (i.e., mtime after the v4.9.0 release date). One-release grace
+  period.
+
+## Migration / coexistence
+
+- Old `snapshot/<entity>.md` files remain readable by readers per the Reader contract step 3.
+- Old `stream/<week>-stream.md` files remain readable similarly.
+- A future v4.10.x `migrate-evidence` skill will optionally rewrite legacy into weekly/ + _index/.
+  Out of scope for v4.9.0.
+- No automatic deletion of legacy folders.
+
+## Cross-references
+
+- `comprehensive-structured-capture.instructions.md` — block shape inside weekly/ files.
+- `meetings-verbatim-required.instructions.md` — raw verbatim/ folder for meetings only.
+- `evidence-layout-canonical.instructions.md` — updated in v4.9.0 to allow weekly/ + _index/.
+- `snapshot-vs-stream.instructions.md` — DEPRECATED in v4.9.0; redirects here.
+- `multi-user-shared-files.instructions.md` — `_index/` is per-contributor; not shared.

package/plugin/instructions/workiq-only.instructions.md

@@ -46,6 +46,143 @@ For every M365 source in scope:
 4. **DO NOT attempt** `m365_get_transcript`, `m365_get_facilitator_notes`, `m365_list_meetings`, `m365_list_events`, `m365_search_files` (for content), `m365_download_file` (for transcript/notes), or any Graph REST URL for the in-scope sources above. These are FORBIDDEN — **including as a "last-resort" fallback after WorkIQ fails** (kushi v4.4.1+). Calling them is a defect; coverage.md must NOT show them in the attempt trail. The deferred-retry marker IS the audit trail for WorkIQ failures.
 5. **Allowed alongside WorkIQ** (structured-data dumps only): `m365_list_chat_messages` for chat-id'd threads → `chat-messages.json`. These run in parallel with the WorkIQ pull, not as a substitute.
 
+## v4.9.0 — Comprehensive Structured Capture (CSC) prompts supersede verbatim prompts
+
+> **v4.9.0:** With the move to Comprehensive Structured Capture (CSC) — see
+> [`comprehensive-structured-capture.instructions.md`](./comprehensive-structured-capture.instructions.md)
+> — pull-* skills now ask WorkIQ for **structured bullet output** instead of raw verbatim bodies.
+> CSC prompts are below in `## CSC canonical prompts (kushi v4.9.0+)`.
+>
+> The **verbatim prompts** in the historical "Canonical WorkIQ commands" section
+> remain valid for exactly one use case: **meetings raw audit folder**
+> (`Evidence/<alias>/meetings/verbatim/<dir>/transcript.txt`) per
+> [`meetings-verbatim-required.instructions.md`](./meetings-verbatim-required.instructions.md).
+> All other sources stop using verbatim prompts as of v4.9.0.
+
+## CSC canonical prompts (kushi v4.9.0+)
+
+Every CSC prompt asks WorkIQ to return one block per entity in CSC shape:
+bulleted sections in canonical order, no prose paragraphs, no omission. Skills
+parse this directly into the `weekly/YYYY-MM-DD_<source>-csc.md` artifact.
+
+The shared closing clause appended to every CSC prompt:
+
+```text
+Return as a Comprehensive Structured Capture (CSC) block per kushi doctrine:
+- Bulleted sections in this exact order: Participants; Topics Discussed; Q&A;
+  Who Said What (material statements only); Decisions; Dates & Numbers Shared;
+  Action Items (with owner and due, mark "owner: unassigned" / "due: unspecified"
+  when not stated); Next Steps; Open Questions; Risks/Blockers/Dependencies;
+  Customer Asks; Artifacts/Links; Coverage Notes.
+- Bullets only, no prose paragraphs anywhere.
+- No omission: do NOT write "and N more", "key highlights only", "sample of",
+  "N of M shown", or "truncated" anywhere outside Coverage Notes.
+- Empty applicable sections explicit as "_None surfaced._".
+- Inapplicable sections OMITTED entirely.
+- Every fact-bearing bullet inline-cites its source (sender, timestamp, message-id,
+  page wdpartid, file URL, work-item id, etc.).
+- Do NOT fabricate from adjacent evidence. If a section's basis is unavailable,
+  write "Source basis: unavailable" in Coverage Notes.
+```
+
+### Meeting CSC block (per meeting)
+
+```text
+For the Teams meeting titled "<subject>" that occurred on <YYYY-MM-DD>, produce
+a Comprehensive Structured Capture block covering the whole meeting. Use the
+transcript and the Copilot recap as the source. Include every distinct topic
+discussed, every question asked AND answered (in Q&A), every commitment made
+(in Action Items with owner+due), every number/date/dollar/percentage stated
+(in Dates & Numbers Shared), and every customer ask. Curate Who Said What to
+material statements only (positions stated, commitments made, disagreements
+voiced) — NOT an utterance log.
+
+<shared closing clause>
+```
+
+### Email thread CSC block (per conversation)
+
+```text
+For the email thread with conversationId <id> (subject "<subject>") between
+<YYYY-MM-DD> and <YYYY-MM-DD>, produce one Comprehensive Structured Capture
+block covering the entire thread. Topics, Q&A, and Action Items must cover
+every reply in the chain. Participants must list all senders and recipients
+across the thread.
+
+<shared closing clause>
+```
+
+### Teams chat thread CSC block (per chat)
+
+```text
+For the Teams chat with chatId <id> ("<topic>") between <YYYY-MM-DD> and
+<YYYY-MM-DD>, produce one Comprehensive Structured Capture block covering the
+full message-by-message thread. Participants must list every poster.
+Q&A captures every question asked AND answered. Who Said What is curated to
+material statements only (3–10 entries typical per 50-message thread).
+
+<shared closing clause>
+```
+
+(Run `m365_list_chat_messages(chatId)` in parallel for the structured JSON
+dump → `chat-messages.json` companion artifact.)
+
+### Teams channel post CSC block (per root message)
+
+```text
+For the Teams channel post in team <teamId> / channel <channelId> with root
+message id <messageId>, produce a Comprehensive Structured Capture block
+covering the post and all replies in the thread.
+
+<shared closing clause>
+```
+
+### OneNote page CSC block (per page — ONE page per call)
+
+```text
+For OneNote page "<title>" (wdpartid <id>) in section "<section>.one", produce
+one Comprehensive Structured Capture block covering the entire page body.
+Topics must cover every heading and section of the page. Dates & Numbers Shared
+must include every number, date, dollar amount, percentage, version, SKU, and
+deadline written on the page. Action Items must list every task/checkbox.
+Artifacts must list every link and embedded file on the page.
+
+If you cannot read the page body, write exactly one line: "body-unavailable"
+and nothing else.
+
+<shared closing clause>
+```
+
+> Tier A (enumeration) and Tier B (search snippets) prompts in the historical
+> section below remain valid — they're used at discovery to enumerate which
+> pages exist. Tier C (per-page) is now this CSC prompt, replacing the
+> v4.8.x "full verbatim body" Tier C prompt.
+
+### SharePoint file CSC block (per file)
+
+```text
+For the SharePoint file at <web URL> (last modified <YYYY-MM-DD>), produce one
+Comprehensive Structured Capture block covering the document. Topics must cover
+every section heading. Dates & Numbers Shared must include every number, date,
+dollar amount, version, deadline. Customer Asks must list every explicit
+customer request in the document. Artifacts must list every link and embedded
+file referenced.
+
+<shared closing clause>
+```
+
+### CRM record CSC block (per opportunity/account/contact)
+
+> CRM is OUT of WorkIQ scope (uses Dataverse REST per `crm-bootstrap-discovery`).
+> But the CSC block shape applies — `pull-crm` constructs the block from REST
+> field reads, not WorkIQ. No WorkIQ prompt needed here; the block schema is in
+> [`comprehensive-structured-capture.instructions.md`](./comprehensive-structured-capture.instructions.md).
+
+### ADO work item CSC block (per work item)
+
+> ADO is OUT of WorkIQ scope (uses ADO REST). Same as CRM: `pull-ado` constructs
+> the CSC block from REST field reads. No WorkIQ prompt needed.
+
 ## Canonical WorkIQ commands (CODIFIED — do not re-discover)
 
 The CLI is resolved in this order: `<workspace>/.kushi/config/user/project-evidence.yml workiq.cli_path` → `Get-Command workiq` (PATH) → `~/.copilot/bin/workiq[.cmd]` (Clawpilot-managed convenience fallback, probed only if it already exists). The legacy `~/.kushi/bin/workiq.cmd` path was removed in v4.4.4 — see the `setup` skill for the functional verification flow.
@@ -198,5 +335,7 @@ If `Result: SUMMARY-ONLY` after doubled-strict retry: the artifact is acceptable
 - `plugin/instructions/crm-bootstrap-discovery.instructions.md` — CRM is OUT of WorkIQ scope; it uses Dataverse REST.
 - `plugin/instructions/ado-bootstrap-discovery.instructions.md` — ADO is OUT of WorkIQ scope; it uses ADO REST.
 - `plugin/instructions/scope-boundaries.instructions.md` — every pull-* still respects integrations.yml#boundaries.
-- `plugin/instructions/evidence-thoroughness.instructions.md` — verbatim is the bar; this rule operationalizes how to hit it.
+- `plugin/instructions/evidence-thoroughness.instructions.md` — CSC bullet thresholds are the depth bar (v4.9.0); this rule operationalizes how to source the bullets.
+- `plugin/instructions/comprehensive-structured-capture.instructions.md` (v4.9.0+) — CSC block schema.
+- `plugin/instructions/weekly-csc.instructions.md` (v4.9.0+) — weekly/ layout + `_index/`.
 - `plugin/instructions/deferred-retry-on-workiq-fail.instructions.md` (kushi v4.4.1+) — defines exactly what happens when WorkIQ fails. NO Graph fallback. Marker + continue + inform.

package/plugin/skills/ask-project/SKILL.md

@@ -1,11 +1,27 @@
 ---
 name: "ask-project"
-version: "2.0.0"
-description: "Read-only Q&A over an already-bootstrapped project. Answers natural-language questions about a project (status, decisions, MACC, stakeholders, risks, action items, timeline, specific docs/meetings) using only the project's State/, Evidence/, and snapshot/ files. Cited; no source pulls; no outbound."
+version: "3.0.0"
+description: "Read-only Q&A over an already-bootstrapped project (v4.9.0+). Answers natural-language questions using State/, Evidence/, and weekly CSC files via 3-step reader fallback (_index/entities.yml → weekly/*.md → legacy snapshot/ + stream/). Cited; no source pulls; no outbound."
 ---
 
 # Skill: ask-project
 
+## v4.9.0 — Reader fallback chain (HARD RULE; per `weekly-csc.instructions.md`)
+
+For every `<alias>/<source>/` accessed:
+
+1. **Primary**: read `Evidence/<alias>/<source>/_index/entities.yml`. For each entity needed, pull `latest_csc_file` + `latest_csc_block_offset`; open the weekly file and read the CSC block by entity anchor.
+2. **Secondary (no _index/, but weekly/ exists)**: walk `Evidence/<alias>/<source>/weekly/*.md` newest-first; parse `## <name> {#<anchor>}` headings; build an in-memory entity index for this query.
+3. **Legacy fallback (pre-v4.9.0)**: if `weekly/` is empty BUT `snapshot/` or `stream/` exist, read those. Cite as legacy.
+
+Citation form in this skill's output:
+- **New (v4.9.0)**: `[source: <alias>/<source>/weekly/<YYYY-MM-DD>_<source>-csc.md#<entity-anchor> · <iso-ts>]`
+- **Legacy (reading from snapshot/+stream/)**: append ` (legacy pre-v4.9.0 layout)` after the citation.
+
+Mark the answer / output footer with `Source-layout: weekly-csc | weekly-csc-plus-legacy | legacy-only` so the user knows whether a refresh would update the layout.
+
+Readers MUST NOT delete or modify legacy folders. Migration is out of scope for v4.9.0.
+
 Run this whenever the user asks a question about a known project — even without an explicit verb.
 
 The user does NOT need to type `/ask-project` or `@Kushi ask`. If the message:
@@ -59,7 +75,7 @@ Map the question to one or more **answer kinds**, which determines which files t
 | "what risks / blockers / dependencies …" | `State/06_risks-and-issues.md` |
 | "what's the timeline / when is …", "milestone for …" | `State/07_timeline-and-milestones.md` |
 | "what's the MACC / ECIF / commercials / $ size for …" | grep `MACC`, `ECIF`, `\$[0-9]`, `consumption commitment` across `State/` + `Evidence/<alias>/_Consolidated/` |
-| "what does the deck / SOW / doc say about …" | `<project>/SharePoint/snapshot/files/*.md` matching the doc name; fallback to `tree.md` to locate |
+| "what does the deck / SOW / doc say about …" | `Evidence/<alias>/sharepoint/weekly/<YYYY-MM-DD>_sharepoint-csc.md#<entity-anchor>` (v4.9.0); fallback: `<project>/SharePoint/snapshot/files/*.md (legacy pre-v4.9.0 layout)` matching the doc name; fallback to `tree.md` to locate |
 | "what was in the email from X about Y" | grep across `Evidence/<alias>/Email/` |
 | "what's in OneNote about …" | `Evidence/<alias>/OneNote/` + `external-context/*onenote*.md` |
 | "summary of the project" | `State/00_overview.md` (this file IS the answer) |
@@ -163,6 +179,10 @@ Explicit triggers also accepted:
 > User: "what did Mark say in last week's call about pricing?"
 > → If no Meetings file for that week: "I don't have a meetings transcript for week of YYYY-MM-DD. Want me to pull meetings for AGCO since YYYY-MM-DD?"
 
-## References (v4.4.7)
-
-- Name → ID resolution (any source) follows `..\..\instructions\fuzzy-disambiguation.instructions.md`.
+## References (v4.4.7)
+
+- Name → ID resolution (any source) follows `..\..\instructions\fuzzy-disambiguation.instructions.md`.
+
+## Changelog
+
+- **v3.0.0 (kushi v4.9.0, 2026-05-26)**: 3-step reader fallback chain (`_index/entities.yml` → `weekly/*.md` → legacy `snapshot/` + `stream/`). New citation form `weekly/<YYYY-MM-DD>_<source>-csc.md#<anchor>`. Legacy citations suffixed `(legacy pre-v4.9.0 layout)`. Output marked with `Source-layout:` footer.

package/plugin/skills/bootstrap-project/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: "bootstrap-project"
-version: "2.4.0"
-description: "First-time setup for a project: machine preflight, side-by-side config, engagement-root + project resolution, customer-hint discovery sweep across all WorkIQ-driven sources (kushi v4.8.0+, per customer-hint-discovery.instructions.md), initial 30d snapshot+stream pull across all enabled sources. Verbatim-by-default per `verbatim-by-default.instructions.md` — every enabled source dispatched, no silent skips. Discovery sweep MANDATORY before declaring blocked-config for email/teams/meetings/sharepoint per customer-hint-discovery.instructions.md. CRM bootstrap discovery REQUIRED per `crm-bootstrap-discovery.instructions.md` (v2.3.0). Writes per-user refresh report per `run-reports.instructions.md`. Cleans stale no-match notes on resolution per `cleanup-on-resolution.instructions.md`. Builds State/ only on `full` profile. Idempotent."
+version: "3.0.0"
+description: "First-time setup for a project (kushi v4.9.0+): machine preflight, side-by-side config, engagement-root + project resolution, customer-hint discovery sweep across all WorkIQ-driven sources, initial full-window CSC pull across all enabled sources writing weekly/ + _index/ per source. All pull-* skills write CSC blocks per comprehensive-structured-capture.instructions.md. Verbatim-by-default — every enabled source dispatched, no silent skips. Discovery sweep MANDATORY before declaring blocked-config. CRM bootstrap discovery REQUIRED per `crm-bootstrap-discovery.instructions.md`. Writes per-user refresh report per `run-reports.instructions.md`. Cleans stale no-match notes on resolution per `cleanup-on-resolution.instructions.md`. Builds State/ only on `full` profile. Idempotent."
 ---
 
 # Skill: bootstrap-project
@@ -18,6 +18,12 @@ description: "First-time setup for a project: machine preflight, side-by-side co
 >
 > **Canonical evidence layout** (HARD, kushi v3.12.1+): every per-source artifact lands under `<project>/Evidence/<alias>/<source>/{snapshot,stream,...}/` and **nowhere else** under `<project>/`. Sibling folders like `<project>/email-context/`, `<project>/notes/`, `<project>/_Weekly Summaries/` are FORBIDDEN. Bootstrap MUST (a) scaffold the canonical tree, and (b) before pulling, scan `<project>/` for non-canonical source-output folders and migrate each to `Evidence/<alias>/<source>/_legacy_<original-name>_pre-bootstrap/` (move, do not delete; log under `## Layout migrations` in the bootstrap report; add a `runs:` entry with `type: layout-migration`). See `evidence-layout-canonical.instructions.md`.
 
+> **v4.9.0 update — CSC + weekly/ layout**:
+> - Drop the v4.8 "1 verbatim transcript per source per bootstrap" carve-out. Bootstrap now walks the FULL boundary window (`boundaries.<source>.date_window_days` or default lookback_days from config) and writes one `weekly/<YYYY-MM-DD>_<source>-csc.md` file per ISO week with activity.
+> - Preflight per source: ensure `<engagement-root>/<project>/Evidence/<alias>/<source>/{weekly,_index}/` directories exist. (`verbatim/` is created ONLY for meetings, by pull-meetings Half A.)
+> - Bootstrap-status.md template language MUST mention `weekly/` + CSC, not snapshot/+stream/.
+> - All pull-* skills invoked by bootstrap now write CSC blocks per `comprehensive-structured-capture.instructions.md`. No AI Narrative Summary section is required.
+
 Run this when the user says any of: "bootstrap a new project", "set up project evidence for `<X>`", "add me to project `<X>`", "I'm new to `<X>`", "do it all for `<X>`" (and the project has no Evidence/ folder yet).
 
 ## Profile-aware behavior
@@ -86,7 +92,7 @@ Optional (only if user enables CRM/ADO — both go into `<workspace>/.kushi/conf
 
 | Live destination | Source |
 |---|---|
-| `<workspace>/.kushi/config/shared/integrations.yml` `crm:` block | already seeded by `templates/init/integrations.template.yml` — fill in `crm.environmentUrl` + `crm.tenantId` |
+| `<workspace>/.kushi/config/shared/integrations.yml` `crm:` block | seeded **PRE-FILLED with Microsoft IS defaults (kushi v4.8.4+)** via `templates/init/integrations.template.yml` — `environment_url: https://iscrm.crm.dynamics.com`, `tenant_id: 72f988bf-...`, `default_entity_set: new_frontierengineeringtriages`. Override only if you work in a non-Microsoft tenant. |
 | `<workspace>/.kushi/config/shared/integrations.yml` `ado:` block | already seeded by `templates/init/integrations.template.yml` — fill in `ado.organization` + `ado.project` |
 
 **Boundaries upgrade for existing projects (v3.7.0+):** If `<engagement-root>/<project>/integrations.yml` already exists but has no top-level `boundaries:` key, append the scaffolded `boundaries:` block from `templates/init/project-integrations.template.yml` (preserving every existing key). Then proceed to Step 4 to populate it.
@@ -155,7 +161,7 @@ After all sweeps complete, write the `## Discovery Sweep Results` table to `<pro
 
 **Boundaries gate** (kushi v3.7.0+, per `scope-boundaries.instructions.md`): before dispatching any `pull-*`, read `<engagement-root>/<project>/integrations.yml#boundaries` and verify each enabled source has its required boundary key populated. After Step 3.5, many of these should now contain discovered rows (annotated `needs_review: true` in their sidecars). For sources where the sweep ran but returned 0 candidates, the status is `unresolved` (not `blocked-config`) — add a one-liner to `<project>/OPEN-QUESTIONS-DRAFT.md` asking the user to widen the hint or manually seed the boundary. For sources where the sweep COULD NOT run because a prerequisite is genuinely missing (CRM/ADO shared connection block empty; SharePoint local-folder discovery), `blocked-config` is correct and the `next_step` MUST cite the specific missing field.
 
-For CRM and ADO additionally verify the shared connection block exists in `<workspace>/.kushi/config/shared/integrations.yml` (`crm:` block with `environmentUrl` + `tenantId`, OR `ado:` block with `organization` + `project`) with non-placeholder values; if missing, prompt the user to fill those two/four fields directly (no separate template files — they live in `templates/init/integrations.template.yml`) and park in Open Questions with the path. **Do NOT auto-improvise** by inferring a tenant/org or by narrating CRM evidence from email — both are explicit anti-patterns in v3.7.0.
+For CRM and ADO additionally verify the shared connection block exists in `<workspace>/.kushi/config/shared/integrations.yml` (`crm:` block with `environment_url` + `tenant_id`, OR `ado:` block with `organization` + `project`) with non-placeholder values. **As of kushi v4.8.4, these blocks ship PRE-FILLED with the canonical Microsoft Industry Solutions defaults** (iscrm.crm.dynamics.com / Microsoft tenant / IndustrySolutions ADO org / IS Engagements project), so on a fresh install this preflight passes by default. Only if a contributor has hand-edited the file to empty values or works in a non-Microsoft tenant will this preflight fail — in that case, prompt the user to fix the affected field directly and park in Open Questions with the path. **Do NOT auto-improvise** by inferring a tenant/org or by narrating CRM evidence from email — both are explicit anti-patterns in v3.7.0.
 
 **CRM discovery is REQUIRED before declaring `disabled` (kushi v3.11.0+, per `crm-bootstrap-discovery.instructions.md`).** If `<workspace>/.kushi/config/shared/integrations.yml` exists and `az` auth succeeds, bootstrap MUST run the full 4-step Dataverse REST resolution sequence from `pull-crm/SKILL.md#resolution-order-when-crmrecordid-is-unset` (title-first → all matching accounts → wide-text → recent-slice → ask user) against the live endpoint before writing `boundaries.crm.disabled: true`. Any other path that sets `disabled: true` is a defect. If steps 1–4 all return 0, present the top 5 candidates from step 4 to the user before final disposition. Log the full attempt trail (queries + counts + outcome) to the bootstrap refresh-report under `## CRM resolution attempts`. If auth fails or Dataverse is unreachable, leave the boundary empty with `reason: 'crm-auth-unavailable-<date>'` — NOT `disabled: true` — so the next refresh retries.
 
@@ -242,7 +248,9 @@ Source order (deterministic, do not reshuffle):
 8. `pull-ado` (if enabled)
 9. `pull-misc` (if `<project>/external-links.txt` exists with ≥ 1 non-placeholder, non-delegated link)
 
-Each produces snapshot/ + stream/ output per `snapshot-vs-stream.instructions.md`. The gate enforces: shape (snapshot/ + stream/ + verbatim/ where applicable) → per-source extras (transcript-class file for meetings, sectionId for onenote, siteId for sharepoint, chatId for teams, folders for email, engagementRecordId for crm, areaPath for ado) → process-compliance (workiq-first / verbatim-by-default / capture-learnings / run-report mention / fuzzy-disambiguation cited / cleanup-on-resolution / citation-ledger).
+> **v4.9.0 SUPERSEDED.** Bootstrap walks the full window now; pull-* write `weekly/<YYYY-MM-DD>_<source>-csc.md` + `_index/entities.yml` per source (not snapshot/ + stream/). See `weekly-csc.instructions.md` and `comprehensive-structured-capture.instructions.md`.
+
+Each produces CSC weekly output per `weekly-csc.instructions.md` + `comprehensive-structured-capture.instructions.md`. The gate enforces: shape (weekly/ + _index/ + verbatim/ for meetings) → per-source extras (transcript-class file for meetings, sectionId for onenote, siteId for sharepoint, chatId for teams, folders for email, engagementRecordId for crm, areaPath for ado) → process-compliance (workiq-first / verbatim-by-default / capture-learnings / run-report mention / fuzzy-disambiguation cited / cleanup-on-resolution / citation-ledger).
 
 **pull-misc bootstrap note:** if `<project>/external-links.txt` does NOT exist, scaffold it from `templates/init/external-links.template.txt` so the user has a place to paste links. Mark the source as `enabled: true, links: 0` in `integrations.yml#boundaries.misc` and skip the dispatch (nothing to fetch yet).
 
@@ -286,3 +294,7 @@ End with a one-liner pointing at Q&A:
 ## Issue Recovery
 
 When this skill exposes a reusable defect (auth pattern, doctrine gap, layout mismatch), apply the [Issue Recovery Rule](../../instructions/issue-recovery.instructions.md): fix the smallest correct repo-owned artifact first, prefer durable fixes over per-run workarounds, then re-run the narrowest failed check. Do NOT use memory as a substitute for correcting the workflow surface.
+
+## Changelog
+
+- **v3.0.0 (kushi v4.9.0, 2026-05-26)**: Drops "1 verbatim per source per bootstrap" carve-out. Walks the full boundary window per source. Preflight ensures `weekly/` + `_index/` exist per source. All pull-* invoked write CSC blocks per comprehensive-structured-capture.instructions.md.