docs(skills): after a compaction, a hold covers only its own work

claude-compaction-restore's three hold-first lines now say a hold you verified as live, a release order or an operator's restore map governs the work it covers, and the restored seat does the unaffected work. The read-depth table and the reconnect note are written as the seat resumes, without holding unaffected work for them.
docs: relay binding with its handle, and what skill restore keeps (#3096 )
2026-11-12 16:23:00 -10:00 · 2026-11-13 00:26:35 +00:00 · 2026-11-12 16:19:04 -08:00
3 changed files with 12 additions and 8 deletions
@@ -163,8 +163,8 @@ is not a restore request.
 1. **Check for a hold first.** Read the restore request, the per-seat instruction file
   (`<OPENRIG_HOME>/compaction/post-compact-extra/<session>.md`, named after your full session, such as
   `dev-impl@my-rig` with `.md` added) when it exists, the newest row or message from whoever routes your work, and any hold
-   from the authority above them. A hold, a release order or an operator's own restore map overrides the default
-   order below. Before any write, also run `rig whoami --json` and `rig queue whoami`.
+   from the authority above them. A hold you verified as live, a release order or an operator's own restore map overrides
+   the default order below for the work it covers; do the unaffected work. Before any write, also run `rig whoami --json` and `rig queue whoami`.
 2. **Name your class and state its read budget** before reading (see "Two restore classes"), with a
   checkpoint at each step below. The budget exists so that the restore leaves room for the work it was
   restored to do. Step 5 takes a high-context seat to its Tier 2 line; the extra sources under "Two restore classes" apply only
@@ -218,7 +218,7 @@ The audit message asks for a read-depth table and tells you not to conserve toke
   - where you are in time: what earlier windows established, what the last window did, what is true now;
   - the connections you have rebuilt, in a few lines;
   - the connections you could not rebuild, and where you would look;
-   - the next authorized step and who authorizes it. If a hold stands, the next step is waiting.
+   - the next authorized step and who authorizes it. If a hold you verified as live covers it, wait on that; do the unaffected work.

 The next restore map links this note, which keeps the chain unbroken.

@@ -238,7 +238,7 @@ unaffected work.
 - Do not launch a fresh session in place of restoring.
 - An honest `PARTIAL` with its reason is a correct outcome. Claiming coverage you did not reach is the
  failure.
- Do not resume task work until the read-depth table and the reconnect note exist.
+- Write the read-depth table and the reconnect note as you resume; don't hold unaffected work for them.

 ## Failure modes

@@ -1278,7 +1278,8 @@ reaching them is ordinary work rather than an escalation.
 - **The `relay` gateway connector** carries a registered person's questions, updates and notices to a
  bridge process you run, over loopback HTTP, and takes their replies back with the sender resolved
  against their bindings. It's off until a person has a relay binding (`rig gateway human add <id> …
-  --binding "relay:<connectorRef>:file:<absolute env file>:primary"`), and the bridge, with its reach to
+  --binding "relay:<connectorRef>:file:<absolute env file>:primary:handle=<alias>"`). The handle is how
+  their replies are matched to them: a binding without one is outbound-only. The bridge, with its reach to
  any messaging platform, is yours (`docs/reference/gateway-relay.md`).
 - **`rig gateway channel add <id> --display-name <name> --binding
  slack:<connectorRef>:<secretsRef>:<channelId>`** — a Slack shared channel as one participant,
@@ -1597,8 +1598,11 @@ your circumstances is configuration, and the ones that are not, another agent ca
  without that storage), so don't copy them by hand; read the `plugin_skill_*` warnings if one
  is missing. TDD is no longer a default skill.
 - **`on_restore: follow_source`** on a spec's `resources.skills[]` entry refreshes that skill from its source when
-  a Claude or Codex seat is restored, keeping its conversation; `keep_saved`, the default, keeps the saved copy, but a
-  skill whose source changed after `rig down` comes back missing, named in the restore's warnings.
+  a Claude or Codex seat is restored, keeping its conversation, while the seat's copy is still what OpenRig wrote: an
+  edited, unrecorded or linked copy is kept and named, and a source that can't be read leaves the saved copy.
+  `keep_saved`, the default, keeps the saved copy, but a missing file comes back only from a source that still has the
+  bytes OpenRig wrote: a file whose source changed after `rig down` stays missing, named in the restore's warnings,
+  even when the rest of the skill (its `SKILL.md`, say) is unchanged.
 - **`rig startup-proof submit`** — answer an explicitly selected authenticated startup challenge.
  Startup adds no proof exercise by default. Declare `startup_proof` with `authenticated` or
  `none` in the applicable startup layers; see `docs/reference/rig-spec.md#startup-block/startup-proof-selection`.
@@ -69,7 +69,7 @@
      "agent-operated-software/SKILL.md": "ec910bf5f631d7f20872786a565a0437a446fa2d5fd8b52b685979978243a6a8",
      "agent-operated-workflows/SKILL.md": "24e54cbfa979999f523516112670618b2926ae5d42c7aa0846c94db214857ea5",
      "applying-a-permission-policy/SKILL.md": "f7bba8d50b10fc8715f40a57bf82677ca9388c6718c8a162833d9c6daf2afdca",
-      "claude-compaction-restore/SKILL.md": "79a707974824e91d5888c7bc10b152989ff891491f1adecf3e93991e2af55f81",
+      "claude-compaction-restore/SKILL.md": "cd8369a62647ad389a9b0cc91c6f784a0d93e2d2b874bf9570d47c2134647d1c",
      "claude-compaction-restore/reference/SKILL-v1.md": "6af2a889c9c38227fb186897c6e29185f81b215822070b469a6ceeb47734e133",
      "claude-compaction-restore/scripts/precompact-hook.mjs": "712e97c8beb02eb6c2815884a960778a225d8f6238dfe04a9b0547c34eaa131d",
      "claude-compaction-restore/scripts/restore-from-jsonl.mjs": "5584cff4ab36cd843e14c81bd201e4a2f8e498e995abacc4f81ec6c1fc556b10",