fdeops 3.10.0 → 3.11.0

package/mcp/recipes/file.md

@@ -1,10 +1,10 @@
 # Recipe: local file / paste (no source MCP)
 
-**Use when:** you already have a transcript, `.eml`, or export on disk — or you paste into chat.
+**Use when:** you already have a transcript, `.eml`, or export on disk — or you paste into chat. This is the default FDE path.
 
 ## Setup
 
-None beyond the FDEOps sink (`fde` CLI and optionally `fdeops-ingest` MCP).
+None. Bound workspace + `fde ingest` (or `@fde debrief` for short notes).
 
 ## Pull phrase
 
@@ -12,14 +12,14 @@ None beyond the FDEOps sink (`fde` CLI and optionally `fdeops-ingest` MCP).
 @fde stage this transcript into the fieldbook and propose updates
 ```
 
-(or attach / point at a path)
+(or attach / point at a path, or paste and say debrief)
 
 ## Agent steps
 
-1. Bind engagement.
-2. `fde ingest stage --source file --title "<short>" <path>` (or stdin).
-3. `fde ingest propose <id>` → rewrite prefixes → show FDE → on confirm `fde ingest apply`.
+1. Bind engagement (`fde resume`).
+2. Short paste → debrief verb. Long file → `fde ingest stage --source file --title "<short>" <path>`.
+3. Propose → rewrite prefixes → show FDE → on confirm apply.
 
 ## mcp.json
 
-Not required for the source. Optional sink only — see [../fdeops-ingest/README.md](../fdeops-ingest/README.md).
+Not required.

package/mcp/recipes/granola.md

@@ -2,16 +2,19 @@
 
 **Use when:** meeting notes live in Granola (or a similar notes MCP). FDEOps does not ship a Granola server — you add whichever MCP/export path you trust.
 
+Daily path if the notes are already in chat or on disk: paste to `@fde debrief` or [file.md](./file.md). This recipe is only for **pull**.
+
 ## Setup (once)
 
-1. Install / enable a **Granola (or notes) MCP** in Cursor/Claude per that product’s docs.
-2. Add the FDEOps **sink** MCP (`fdeops-ingest`) — [../fdeops-ingest/README.md](../fdeops-ingest/README.md).
-3. Reload MCP / restart the agent host.
-4. Test: `@fde what can you pull?` — agent should see both sink tools and the notes source tools.
+1. Enable a **Granola (or notes) MCP** in Cursor/Claude per that product’s docs.
+2. Reload MCP / restart the host.
+3. Test: `@fde what can you pull?` — notes-source tools should appear.
+
+The sink is **`fde ingest` in this bound workspace.** You do not need `fdeops-ingest` MCP for daily pull.
 
 ### Example mcp.json shape (illustrative)
 
-Replace `granola-mcp` command/args with whatever the real server documents. FDEOps only needs *some* tool that returns transcript text.
+Replace command/args with whatever the real Granola MCP documents. FDEOps only needs *some* tool that returns transcript text.
 
 ```json
 {
@@ -22,19 +25,12 @@ Replace `granola-mcp` command/args with whatever the real server documents. FDEO
       "env": {
         "GRANOLA_API_KEY": "from-your-secrets"
       }
-    },
-    "fdeops-ingest": {
-      "command": "node",
-      "args": ["/absolute/path/to/fdeops/mcp/fdeops-ingest/server.js"],
-      "env": {
-        "FDEOPS_ENGAGEMENT": "/Users/you/fde-engagements/acme/.fde"
-      }
     }
   }
 }
 ```
 
-**No Granola MCP available?** Export transcript to a file → follow [file.md](./file.md).
+**No Granola MCP available?** Export transcript to a file → [file.md](./file.md).
 
 ## Pull phrase
 
@@ -44,15 +40,15 @@ Replace `granola-mcp` command/args with whatever the real server documents. FDEO
 
 ## Agent steps
 
-1. Capability check — if no notes/Granola tools, run connect flow (`ingest-connect.md`).
-2. Fetch transcript via source MCP (or ask which meeting).
-3. `ingest_stage` / `fde ingest stage --source granola --title "…"`.
-4. Propose → confirm → apply. Never auto-apply.
+1. Capability check — notes-source tools present?
+2. Fetch transcript text (ask which meeting if ambiguous).
+3. `fde ingest stage --source granola --title "<short>"` then propose → confirm → apply.
+4. Extract decisions/risks/next — do not dump the raw transcript into `.fde/`.
 
 ## Common fails
 
 | Symptom | Fix |
 |---------|-----|
-| Agent says it can’t reach Granola | MCP not saved / host not reloaded / wrong env key |
-| Wrong client inbox | Set `FDEOPS_ENGAGEMENT` or bind workspace (`fde resume --init`) |
-| Empty propose | Agent must rewrite `.debrief-propose` with type prefixes |
+| No Granola tools | Source MCP not loaded — they save + reload |
+| Wrong meeting | One clarifying question, then fetch |
+| Wrong engagement | `fde resume` in this workspace before staging |

package/mcp/recipes/notion.md

@@ -1,13 +1,14 @@
 # Recipe: Notion docs / meeting notes
 
-**Use when:** useful engagement notes live in Notion. FDEOps does not ship a Notion server — use a Notion MCP (or export markdown).
+**Use when:** useful engagement notes live in Notion. FDEOps does not ship a Notion server — use a Notion MCP (or export markdown). We do not write back to Notion.
 
 ## Setup (once)
 
-1. Enable a **Notion MCP** (official or community) with a token that can read the pages you need.
-2. Add **fdeops-ingest** sink — [../fdeops-ingest/README.md](../fdeops-ingest/README.md).
-3. Reload MCP / restart host.
-4. Test: `@fde what can you pull?`
+1. Enable a **Notion MCP** (official or community) with a token that can **read** the pages you need.
+2. Reload MCP / restart host.
+3. Test: `@fde what can you pull?`
+
+The sink is **`fde ingest` in this bound workspace.** `fdeops-ingest` MCP is optional.
 
 ### Example mcp.json shape (illustrative)
 
@@ -20,13 +21,6 @@
       "env": {
         "NOTION_TOKEN": "from-your-secrets"
       }
-    },
-    "fdeops-ingest": {
-      "command": "node",
-      "args": ["/absolute/path/to/fdeops/mcp/fdeops-ingest/server.js"],
-      "env": {
-        "FDEOPS_ENGAGEMENT": "/Users/you/fde-engagements/acme/.fde"
-      }
     }
   }
 }
@@ -42,15 +36,19 @@
 
 ## Agent steps
 
-1. Capability check — Notion tools present?
-2. Fetch page/block text via Notion MCP (ask which page if ambiguous).
-3. Stage with `--source notion`.
-4. Propose → confirm → apply.
+1. Capability check — Notion read tools present?
+2. Fetch page/block text (ask which page if ambiguous).
+3. `fde ingest stage --source notion --title "<short>"` → propose → confirm → apply.
+
+## Never
+
+- Create or edit Notion pages from FDEOps.
+- Ambient-sync a database.
 
 ## Common fails
 
 | Symptom | Fix |
 |---------|-----|
 | 401 / forbidden | Token lacks access to that workspace/page |
-| Huge page dump | Stage full text in `.inbox/`; propose only short dated facts |
-| Wrong engagement | Bind / `FDEOPS_ENGAGEMENT` |
+| Huge page dump | Stage full text in `.inbox/`; apply only short dated facts |
+| Wrong engagement | Bind this workspace before staging |

package/mcp/recipes/slack.md

@@ -0,0 +1,61 @@
+# Recipe: Slack (pull only)
+
+**Use when:** useful context lives in a Slack thread or channel. FDEOps does **not** ship a Slack server, does **not** post messages, and does **not** sync channels.
+
+You add whatever Slack MCP your host already supports. We only accept **text you pulled**, then the same confirm loop as a debrief.
+
+## Setup (once)
+
+1. Enable a **Slack MCP** (official or community) with read access to the threads you need.
+2. Reload MCP / restart the host.
+3. Test: `@fde what can you pull?` — Slack fetch tools should appear. The FDEOps **CLI** (`fde ingest`) is the sink if this workspace is bound; you do not need `fdeops-ingest` MCP for daily use.
+
+### Example mcp.json shape (illustrative)
+
+Replace the Slack server with whatever that MCP actually documents. Do not invent a package name.
+
+```json
+{
+  "mcpServers": {
+    "slack": {
+      "command": "npx",
+      "args": ["-y", "YOUR-SLACK-MCP-PACKAGE"],
+      "env": {
+        "SLACK_BOT_TOKEN": "from-your-secrets"
+      }
+    }
+  }
+}
+```
+
+Optional sink MCP (only if you are not running `fde ingest` from this workspace): see [../fdeops-ingest/README.md](../fdeops-ingest/README.md). Pass `engagement` as the path to this client's `.fde/` (from `fde resume --bind`).
+
+**No Slack MCP?** Copy the thread to a file or paste into chat → [file.md](./file.md) or `@fde debrief`.
+
+## Pull phrase
+
+```text
+@fde pull yesterday's #acme-launch thread into the fieldbook
+```
+
+## Agent steps
+
+1. Capability check — Slack **read** tools present? If not → this recipe, then stop.
+2. Fetch the thread/channel as **text** (ask which channel/thread if ambiguous).
+3. `fde ingest stage --source slack --title "<short>"` (CLI in this bound workspace).
+4. Propose → FDE confirms → apply. Extract decisions/risks/asks — do not dump the thread into `.fde/`.
+
+## Never
+
+- Post, reply, or react in Slack from FDEOps.
+- Background-sync a channel.
+- Auto-apply.
+
+## Common fails
+
+| Symptom | Fix |
+|---------|-----|
+| No Slack tools | Source MCP not loaded — they save + reload; we cannot silent-install |
+| Missing channel | Token/scopes cannot read that workspace — their Slack admin, not FDEOps |
+| Huge dump | Stage full text in `.inbox/`; apply only short dated facts |
+| Wrong client | Bind this workspace (`fde resume`) before staging |

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "fdeops",
-  "version": "3.10.0",
+  "version": "3.11.0",
   "description": "Field kit for engineers embedded in client work - a real CLI (recon, memory, portfolio), one @fde skill with field judgment on top, and hooks that make it automatic. Claude Code plugin and any agent that loads skills.",
   "bin": {
     "fdeops": "bin/install.js",

package/plugin.json

@@ -1,7 +1,7 @@
 {
   "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
   "name": "fdeops",
-  "version": "3.10.0",
+  "version": "3.11.0",
   "description": "Engagement fieldbook for Forward Deployed Engineers: per-client memory in local .fde/ files, one @fde skill, land to close methodology. Local-only, no network.",
   "author": {
     "name": "Subash Natarajan",

package/skills/fde/SKILL.md

@@ -83,13 +83,14 @@ When NOT to interrogate or challenge: unambiguous one-liners, mechanical ops, FD
 | (session entry / where are we) | `fde resume` or use injected TRIAGE; `fde resume --init <name>` only if unbound |
 | Day-1 look at the repo | `fde scan` - then you interpret against the brief |
 | "Debrief these notes" / pastes meeting notes | Prefer `fde debrief --smart <notes>` → **you** (the agent) rewrite `.debrief-propose` with `decision:`/`risk:`/`delivery:`/`contact:`/`next:` prefixes where needed → show FDE → on confirm `fde debrief --apply`. `--smart` is a prefix/keyword gate, not a brain. Fallback: structure prefixed lines yourself, show FDE, then `fde debrief` |
-| "Make sure we're up to date" / "pull relevant info" / "pull from Granola/email/transcript" | Bind engagement; **capability check** (which source MCPs exist this session — never pretend). If missing → connect flow. Else fetch → `fde ingest stage` → `fde ingest propose` → rewrite prefixes → show FDE → on confirm `fde ingest apply`. **Never auto-apply. Never ambient sync.** Detail: `references/ingest.md` |
-| "Connect a new MCP" / "connect Granola/Notion" / "what can you pull?" | Follow `references/ingest-connect.md`: ask which source → emit `mcp.json` from `mcp/recipes/` + sink block → they save/reload in Cursor/Claude → verify tools → optional test stage to `.inbox/` only. You cannot silently install host MCPs. |
+| "Make sure we're up to date" / "pull relevant info" / "pull from Granola/Slack/transcript" | Bind engagement; **capability check** (which *source* MCPs exist — never pretend). If missing → connect flow. Else fetch text → `fde ingest stage` **in this workspace** → propose → rewrite prefixes → show FDE → on confirm `fde ingest apply`. MCP sink is optional; if used, pass `engagement` from `fde resume --bind`. **Never auto-apply. Never push. Never ambient sync.** Detail: `references/ingest.md` |
+| "Connect a new MCP" / "connect Granola/Slack/Notion" / "what can you pull?" | Follow `references/ingest-connect.md`: source MCP only; sink is `fde ingest` here. They save/reload; you cannot silent-install. Paste still works with no MCP. |
 | "Prep me for the meeting with …" / walk-in brief | `fde prep "<short label>"` - present the brief in plain language; do not invent facts missing from `.fde/` |
 | "When did we agree…?" / scope dispute | `fde receipts <term>` - answer with dates; no hit = gap, not proof |
 | "Draft the sponsor update" / how are we doing | `fde status` then follow `references/status.md` for the narrative |
 | "Log that they went quiet" / trust signal | `fde log contact "…" --signal amber\|green\|red`. If they already named the color ("log that as amber"), that is the confirm — write it. If they only described the situation, playback the color once, then write. |
 | Want the HTML fieldbook | `fde dashboard` |
+| "Open my clients in Obsidian" / one window over everything / "can I show this to the sponsor?" | `fde vault` (add `--redacted` for a shared screen). Derived and disposable: it is rebuilt from `.fde/` on every run and never read back, so tell them to keep logging to the fieldbook, not to the vault. |
 | "Clean up the fieldbook" / hygiene / memory feels messy | `fde doctor` - walk issues in plain language; propose fixes; never auto-rewrite without confirm. Includes structural gaps: empty operating map (plan+), stakeholder name forks (Denise vs Denise Chen), duplicates, ship/close risks. Contradictions need judgment (brief vs reality) - doctor is structural; you handle meaning. |
 | "Scrub this secret / redact that token" (buried line, not just last write) | `fde redact <term>` preview, then `fde redact <term> --apply` after confirm. Undo is last-write only; redact is for buried lines. Remind them to rotate the real credential. |
 
@@ -271,8 +272,8 @@ Running the engagement and ending it well.
 | Weekly update due, "need to send the sponsor something" | status | `references/status.md` |
 | Demo coming up, show-and-tell, exec walkthrough | demo-prep | `references/demo-prep.md` |
 | Just out of a meeting, raw notes, "they said…", "debrief" | debrief | the debrief verb (above) + `references/debrief.md` |
-| Make sure we're up to date, pull what's relevant, fetch from Granola/Gmail/transcript | ingest | `references/ingest.md` (capability check → stage → propose → confirm → apply) |
-| Connect a new MCP / connect Granola or Notion / what can you pull | ingest-connect | `references/ingest-connect.md` (+ `mcp/recipes/`) |
+| Make sure we're up to date, pull what's relevant, fetch from Granola/Slack/Gmail/transcript | ingest | `references/ingest.md` (capability check → stage → propose → confirm → apply) |
+| Connect a new MCP / connect Granola Slack or Notion / what can you pull | ingest-connect | `references/ingest-connect.md` (+ `mcp/recipes/`) |
 | Prep me for a meeting / walk-in brief / "what should I know before I talk to…" | - | run `fde prep "<label>"`, present in plain language |
 | Sponsor's boss needs a summary, board update, justify continued investment | exec-narrative | `references/exec-narrative.md` |
 | Status across all my customers | dashboard | `references/dashboard.md` |

package/skills/fde/references/ingest-connect.md

@@ -1,36 +1,34 @@
 # ingest-connect - wire a source MCP in plain language
 
-**Enter when:** the FDE says "I want to connect a new MCP", "connect Granola / Notion / Drive", "how do I pull from …", or a pull request fails because no source tools exist.
+**Enter when:** the FDE says "I want to connect a new MCP", "connect Granola / Slack / Notion", "how do I pull from …", or a pull request fails because no source tools exist.
 
-**Read first:** `references/ingest.md` (sink contract). Recipe pack: `mcp/recipes/` in the fdeops install (file, granola, notion).
+**Read first:** `references/ingest.md` (sink contract). Recipes: `mcp/recipes/` (file, granola, slack, notion).
 
 **Who runs setup:** you guide; the **host** (Cursor/Claude) must save MCP config. You cannot silently install servers into the host.
 
 ## Honest contract
 
-- FDEOps = **sink** (`fdeops-ingest` / `fde ingest`). Sources = **whatever MCP the FDE adds**.
-- You produce a ready config snippet + steps. They save + reload. Then you verify with a capability check + optional test stage.
-- Never invent that Granola/Notion is available if tools are missing. Never ambient sync. Never auto-apply to `.fde/`.
+- **Daily work does not need a source MCP.** Paste notes → debrief. File on disk → `fde ingest`.
+- **Connect means a source**, not FDEOps. Granola/Slack/Notion credentials stay with that MCP. FDEOps never pushes, never ambient-syncs, never stores their tokens.
+- **Sink is the CLI in this bound workspace** (`fde ingest`). `fdeops-ingest` MCP is optional. If you use it, pass `engagement` as the `.fde/` path from `fde resume --bind` (MCP servers often do not inherit the workspace bind).
+- Never invent that Granola/Slack is available if tools are missing. Never auto-apply to `.fde/`.
 
 ## Method
 
-1. **Ask one question** — which source? (`file` / `granola` / `notion` / other name). If "other", ask for the MCP package or docs URL they intend to use.
+1. **Ask one question** — which source? (`file` / `granola` / `slack` / `notion` / other). If "other", ask for the MCP they intend to use. If they just want paste → send them to debrief and stop.
 2. **Capability check (current session)** — list MCP tools you can actually call:
-   - Sink present? (`ingest_stage` / `ingest_list` / or `fde ingest` CLI)
-   - Source present? (anything that can fetch that system's content)
+   - Sink: `fde ingest` CLI (preferred) and/or `ingest_stage`
+   - Source: anything that can **fetch** that system's content (not post)
    - Say clearly: *available now* vs *needs config*.
-3. **Emit config** — open the matching recipe under `mcp/recipes/<source>.md`. Fill absolute paths:
-   - path to `mcp/fdeops-ingest/server.js` (from this fdeops install)
-   - `FDEOPS_ENGAGEMENT` → this client's `…/<slug>/.fde`
-   - placeholders for source API keys (tell them to paste secrets into host env — do not commit keys into the fieldbook)
-4. **Tell them where to paste** — Cursor: MCP settings / `~/.cursor/mcp.json` (or project MCP). Claude Code: MCP config per their docs. One sentence: save → reload MCP / restart session.
-5. **Verify** — after they confirm reload: re-run capability check. If source tools appear, offer a **test pull** into `.inbox/` only (stage + show list). Stop before apply unless they ask to propose.
-6. **Handoff phrase** — give them the daily line, e.g. `@fde pull today's Acme Granola into the fieldbook`.
+3. **Emit config for the source only** — open `mcp/recipes/<source>.md`. Fill placeholders from *that product's* docs. Tell them to paste secrets into host env — never into `.fde/`.
+4. **Tell them where to paste** — Cursor MCP settings / `mcp.json`. Claude Code: their MCP config. Save → reload MCP / restart session.
+5. **Verify** — after reload: re-run capability check. If source tools appear, offer a **test pull** staged to `.inbox/` only. Stop before apply unless they ask to propose.
+6. **Handoff phrase** — e.g. `@fde pull today's Acme Granola into the fieldbook`.
 
-## If they only want the sink
+## If they only want paste / files
 
-Still wire `fdeops-ingest` (or rely on CLI). File drops work with [mcp/recipes/file.md](../../../mcp/recipes/file.md) without any source MCP.
+Do not add MCP. Use debrief or [mcp/recipes/file.md](../../../mcp/recipes/file.md).
 
 ## Checkpoint
 
-Before ending connect: (1) sink reachable, (2) source reachable or honest gap, (3) they know the pull phrase. Do not write `.fde/` during connect.
+Before ending connect: (1) they know paste still works, (2) source reachable or honest gap, (3) they know the pull phrase. Do not write `.fde/` during connect.

package/skills/fde/references/ingest.md

@@ -2,7 +2,7 @@
 
 **Enter when:** the FDE wants to catch the engagement up from external sources — "make sure Acme is up to date," "pull what's relevant," "grab today's Granola and Denise's last email." Raw transcripts and long emails that are too big to paste usefully.
 
-**Connect / capability (different entry):** "connect a new MCP", "connect Granola/Notion", "what can you pull?" → `references/ingest-connect.md` first. Recipes: `mcp/recipes/` (file, granola, notion).
+**Connect / capability (different entry):** "connect a new MCP", "connect Granola/Slack/Notion", "what can you pull?" → `references/ingest-connect.md` first. Recipes: `mcp/recipes/` (file, granola, slack, notion).
 
 **Read first:** `context.md` (what's already logged, what's stale). Bind the engagement before staging anything.
 
@@ -11,7 +11,8 @@
 ## Honest contract (read once)
 
 - FDEOps owns the **sink only**: stage raw pulls → propose → confirm → apply. Nothing writes `.fde/` unreviewed.
-- **Source MCPs are the FDE's.** Granola, Gmail, Notion, custom — whatever they configured in Cursor/Claude. fdeops does not bundle OAuth, connectors, or ambient sync.
+- **Source MCPs are the FDE's.** Granola, Slack, Notion, Gmail, custom — whatever they configured in Cursor/Claude. fdeops does not bundle OAuth, connectors, or ambient sync, and **does not push** to those tools.
+- Prefer **`fde ingest` in this bound workspace.** Optional `fdeops-ingest` MCP: pass `engagement` (path to `.fde/` from `fde resume --bind`) because MCP cwd often is not the client workspace.
 - The core `fde` CLI stays local (git + file reads). Source credentials live with that MCP; fdeops never stores them.
 - After apply, raw stays in `.inbox/`; the system of record (`.fde/`) stays thin dated facts.
 
@@ -20,7 +21,7 @@
 List what you can actually call **this session**:
 
 1. **Sink** — `ingest_stage` / `fde ingest` available?
-2. **Sources** — which fetch tools exist (Granola-shaped, Notion, Drive, file-only)?
+2. **Sources** — which fetch tools exist (Granola-shaped, Slack, Notion, Drive, file-only)?
 3. Tell the FDE in one line: *I can pull from X; Y is not connected.* If they asked to pull Y and it is missing → switch to `ingest-connect.md`. Never pretend a source exists.
 
 ## Ground loop (you do this work)
@@ -67,6 +68,6 @@ Before apply, read back the 2–3 most consequential captures in one breath —
 
 ## Principles
 
-- Pull on request, not on a schedule. No auto-poll, no vacuum of inbox or Slack.
+- Pull on request, not on a schedule. No auto-poll, no vacuum of inbox or Slack. No posting back.
 - Staging is not memory. Only `--apply` after confirm writes `.fde/`.
 - Large artifact → ingest stage first; pasted short notes → debrief verb directly (`references/debrief.md`).

package/templates/.fde/assumptions.md

@@ -4,8 +4,8 @@
 
 | # | Assumption | Blast radius | How we test | Status | Evidence |
 |---|------------|--------------|-------------|--------|----------|
-| 1 | *(seed from brief - stated, unverified)* | CRITICAL / LOAD-BEARING / CONVENIENCE | | OPEN | |
 
-**Status values:** `OPEN` · `TESTING` · `CONFIRMED` · `DISPROVED` · `PARKED`
+**Blast radius:** `CRITICAL` · `LOAD-BEARING` · `CONVENIENCE`  
+**Status:** `OPEN` · `TESTING` · `CONFIRMED` · `DISPROVED` · `PARKED`
 
 **Rule:** a CRITICAL assumption still OPEN blocks plan. DISPROVED → update `reality.md` / `success.md` and log the reset in `decisions.md` the same day.

package/templates/.fde/decisions.md

@@ -6,13 +6,5 @@
 
 ## Decision log
 
-<!-- Use this format for each decision. The rationale is the part that survives the engagement. -->
-
-<!--
-### [Date] Decision title
-- Context: what prompted this decision
-- Options considered: what was on the table
-- Decision: what was chosen
-- Rationale: why this over the alternatives
-- Owner: who approved
--->
+<!-- Dated lines: what was chosen, and why over the alternatives. Long form: docs/schema.md
+     - [2026-07-18] kept the vendor connector over an in-house rewrite - Priya wants the Q3 audit clean first -->

package/templates/.fde/delivery.md

@@ -6,7 +6,9 @@
 
 | Date | Slice | Bucket | Promised | Measured | Accepted by | Evidence | Rollback |
 |------|-------|--------|----------|----------|-------------|----------|----------|
-| | | *(cost-save / risk-mitigation / revenue-uplift)* | *(what we said it would change)* | *(what actually changed, or pending)* | *(customer-side name + date, or it stays claimed)* | *(who/when/metric)* | |
+
+**Bucket:** `cost-save` · `risk-mitigation` · `revenue-uplift`  
+**Accepted by:** a customer-side name and date - without one the value stays *claimed*, not accepted.
 
 ## Ship receipts
 
@@ -14,7 +16,9 @@
 
 | Date | Slice | Audit receipt | Eval receipt |
 |------|-------|---------------|--------------|
-| | | *(dated cite: exceptions/operating path verified — terrain/reality/audit)* | *(n/a \| evals.md pass + HITL owner)* |
+
+**Audit receipt:** dated cite that the exception/operating path was verified (`terrain.md` / `reality.md`).  
+**Eval receipt:** `n/a` unless AI touches the slice, else `evals.md` pass + the HITL owner.
 
 ## Shipped
 

package/templates/.fde/evals.md

@@ -13,13 +13,11 @@
 
 | ID | Input (sanitized) | Expected | Pass rule | Last run | Result |
 |----|-------------------|----------|-----------|----------|--------|
-| G1 | | | | | |
 
 ## Failure modes
 
 | Mode | How it shows up | Detection | Mitigation |
 |------|-----------------|-----------|------------|
-| | | | |
 
 ## Pass / fail (this ship)
 - **Golden:** _/_ pass (threshold: _)
@@ -31,7 +29,6 @@
 
 | Decision / action | Autonomous OK? | Reviewer role | Escalation |
 |-------------------|----------------|---------------|------------|
-| | | | |
 
 **HITL rule for this ship:**
 
@@ -39,4 +36,3 @@
 
 | Date | What changed | Pack re-run? | Notes |
 |------|--------------|--------------|-------|
-| | | | |

package/templates/.fde/risks.md

@@ -2,4 +2,6 @@
 
 | Risk | Status | Owner | Mitigation |
 |------|--------|-------|------------|
-| | open | | |
+
+**Status:** `open` · `mitigating` · `closed` · `accepted`  
+<!-- `fde log risk "<text>"` appends a dated line; the table is for live, owned risk. -->

package/templates/.fde/stakeholders.md

@@ -5,7 +5,8 @@
 
 | Name | Role | Stance | Notes |
 |------|------|--------|-------|
-| | | champion / neutral / resistant | |
+
+**Stance:** `champion` · `neutral` · `resistant`
 
 **Trust signal:** green | amber | red  
 **Last trust check:**

package/templates/.fde/terrain.md

@@ -13,7 +13,6 @@
 
 | Exception / break | Who notices first | What they do today (workaround) | System of record then | Blast if wrong | Evidence |
 |-------------------|-------------------|---------------------------------|-----------------------|----------------|----------|
-| | | | | | |
 
 **Shadow systems / silent workarounds:**
 **Sacred / untouchable in ops:**
@@ -23,4 +22,3 @@
 
 | Step | Deterministic | Model judgement | Human approve |
 |------|---------------|-----------------|---------------|
-| | | | |