State When an Issue Belongs in a Platform Handoff Lane

.agents/skills/backlog-burndown/SKILL.md

@@ -130,6 +130,13 @@ stands rather than being assumed cleared. The round's report names every issue i
 and the blocker each one waits on, since the maintainer reads the report rather than the issue
 bodies, and a stuck issue nobody names reads as ordinary backlog that simply never moves.
 
+An issue that needs a platform, by the test `session-handoff` "The Chain" states, is counted and
+is not ranked. It belongs to that platform's lane, so this run gives it no group, no worker, and no
+claim. An issue only part of which needs a platform is held back the same way. It waits on the
+split "The Chain" describes, which this run does not make, and the narrowed issue is ranked once
+that split lands. The round's report names each issue held back this way, its platform, and
+whether it waits on a split, as it names a `blocked` one.
+
 An issue that asks a question rather than states a defect is not ranked and is never guessed at.
 It has no group, no worker, and no claim, so nothing in "Raising a Blocked Question" applies to it
 except how the question travels. It goes to the maintainer at the end of ranking, per

.agents/skills/session-handoff/SKILL.md

@@ -62,6 +62,26 @@ without trusting the parts of it that have gone stale.
 - **One open handoff issue per track per repository.** A track is a short kebab-case slug naming a
   lane of work, and a session that names no track uses `default`. Parallel lanes each name their
   own, so each closes its own predecessor and neither reads the other's state as current.
+- **A platform lane is for work that needs its platform.** A platform track is one kept for a single
+  host platform, such as `wsl` or `windows`. It takes an issue only when its defect reproduces, or
+  its fix can be verified, on that platform alone. It also needs that no Linux host outside WSL can
+  both fix the issue and verify the fix. An issue meeting both goes to its platform's lane. Any
+  other issue goes to the ordinary backlog, since a lane costs a separate chain and a host-bound
+  session. A defect in WSL interop, the `WSLInterop` binfmt handler, `/mnt/<drive>` paths, or
+  `wslpath` reproduces only inside WSL. So does running a Windows `.exe` from Linux, although WSL
+  runs on Windows. Behavior that differs under Windows PowerShell 5.1 or `winget`, or in Windows
+  path and ACL semantics, reproduces only on Windows. Either still goes to the backlog where a Linux
+  host outside WSL verifies its fix, as a plain unit test of a path translation can. An issue in
+  systemd user scopes, apt, a bash script, or a `pwsh` script run in a Linux container gets no lane
+  by its topic. The two conditions decide it like any other issue. Where only part of an issue needs
+  a platform, the session finding it files each such part as its own issue for its platform's lane.
+  The original is narrowed to the rest, or closed in favor of those issues where nothing is left. A
+  session that may not write issues names the unsplit issue in its report or handoff instead. An
+  issue for a platform whose lane was closed out reopens that lane with `new`, which chains onto the
+  closed link. A platform lane is closed out with no successor once none of its next steps needs its
+  platform. Before that, each remaining step with no issue of its own is filed as one. The lane's
+  last link carries the outcome as a comment, as `AGENTS.md` "Session Scope" closes a track whose
+  work is complete.
 - **The title is for humans**, shaped `Session Handoff [<track>]: <subject>`, which `new` composes
   from the track and the subject it is given, so what a session writes is the subject alone.
   Nothing parses the title, so a maintainer is free to rename one.

.agents/skills/unattended-handoff/SKILL.md

@@ -154,7 +154,10 @@ does not qualify, since skipping one costs nothing and a guess costs a revert an
   and no comment asks the maintainer something still unanswered.
 - **The fix stays inside this repository's tree.** It changes no repository setting, ruleset,
   visibility, secret, or release condition, and needs no credential, account, or host the session
-  lacks.
+  lacks. The issue needs no platform, in whole or in part, by the test `session-handoff` "The
+  Chain" states. An issue that wholly needs a platform belongs to that platform's lane whatever host
+  the session runs on. One that partly needs it waits on the split that section describes, a write
+  the picker does not make.
 - **The fix edits nothing under the hub's `host-setup/agent-safety/`.** That tree is the agent's
   own guard, and it is worked only in an attended session, so a human approves every change to it.
 - **It reverses no settled decision** recorded in an issue, a handoff, or the rule text.

.claude-plugin/fleet-skills/.source-digests/backlog-burndown

@@ -1 +1 @@
-a44167992b6a50fe
+f9045d01bca0c44a

.claude-plugin/fleet-skills/.source-digests/session-handoff

@@ -1 +1 @@
-71ea890bf0050608
+2c0a84e0767fc97c

.claude-plugin/fleet-skills/.source-digests/unattended-handoff

@@ -1 +1 @@
-edad0bee58257e24
+06ef313c4d01b168

.claude-plugin/fleet-skills/skills/backlog-burndown/SKILL.md

@@ -130,6 +130,13 @@ stands rather than being assumed cleared. The round's report names every issue i
 and the blocker each one waits on, since the maintainer reads the report rather than the issue
 bodies, and a stuck issue nobody names reads as ordinary backlog that simply never moves.
 
+An issue that needs a platform, by the test `session-handoff` "The Chain" states, is counted and
+is not ranked. It belongs to that platform's lane, so this run gives it no group, no worker, and no
+claim. An issue only part of which needs a platform is held back the same way. It waits on the
+split "The Chain" describes, which this run does not make, and the narrowed issue is ranked once
+that split lands. The round's report names each issue held back this way, its platform, and
+whether it waits on a split, as it names a `blocked` one.
+
 An issue that asks a question rather than states a defect is not ranked and is never guessed at.
 It has no group, no worker, and no claim, so nothing in "Raising a Blocked Question" applies to it
 except how the question travels. It goes to the maintainer at the end of ranking, per

.claude-plugin/fleet-skills/skills/session-handoff/SKILL.md

@@ -62,6 +62,26 @@ without trusting the parts of it that have gone stale.
 - **One open handoff issue per track per repository.** A track is a short kebab-case slug naming a
   lane of work, and a session that names no track uses `default`. Parallel lanes each name their
   own, so each closes its own predecessor and neither reads the other's state as current.
+- **A platform lane is for work that needs its platform.** A platform track is one kept for a single
+  host platform, such as `wsl` or `windows`. It takes an issue only when its defect reproduces, or
+  its fix can be verified, on that platform alone. It also needs that no Linux host outside WSL can
+  both fix the issue and verify the fix. An issue meeting both goes to its platform's lane. Any
+  other issue goes to the ordinary backlog, since a lane costs a separate chain and a host-bound
+  session. A defect in WSL interop, the `WSLInterop` binfmt handler, `/mnt/<drive>` paths, or
+  `wslpath` reproduces only inside WSL. So does running a Windows `.exe` from Linux, although WSL
+  runs on Windows. Behavior that differs under Windows PowerShell 5.1 or `winget`, or in Windows
+  path and ACL semantics, reproduces only on Windows. Either still goes to the backlog where a Linux
+  host outside WSL verifies its fix, as a plain unit test of a path translation can. An issue in
+  systemd user scopes, apt, a bash script, or a `pwsh` script run in a Linux container gets no lane
+  by its topic. The two conditions decide it like any other issue. Where only part of an issue needs
+  a platform, the session finding it files each such part as its own issue for its platform's lane.
+  The original is narrowed to the rest, or closed in favor of those issues where nothing is left. A
+  session that may not write issues names the unsplit issue in its report or handoff instead. An
+  issue for a platform whose lane was closed out reopens that lane with `new`, which chains onto the
+  closed link. A platform lane is closed out with no successor once none of its next steps needs its
+  platform. Before that, each remaining step with no issue of its own is filed as one. The lane's
+  last link carries the outcome as a comment, as `AGENTS.md` "Session Scope" closes a track whose
+  work is complete.
 - **The title is for humans**, shaped `Session Handoff [<track>]: <subject>`, which `new` composes
   from the track and the subject it is given, so what a session writes is the subject alone.
   Nothing parses the title, so a maintainer is free to rename one.

.claude-plugin/fleet-skills/skills/unattended-handoff/SKILL.md

@@ -154,7 +154,10 @@ does not qualify, since skipping one costs nothing and a guess costs a revert an
   and no comment asks the maintainer something still unanswered.
 - **The fix stays inside this repository's tree.** It changes no repository setting, ruleset,
   visibility, secret, or release condition, and needs no credential, account, or host the session
-  lacks.
+  lacks. The issue needs no platform, in whole or in part, by the test `session-handoff` "The
+  Chain" states. An issue that wholly needs a platform belongs to that platform's lane whatever host
+  the session runs on. One that partly needs it waits on the split that section describes, a write
+  the picker does not make.
 - **The fix edits nothing under the hub's `host-setup/agent-safety/`.** That tree is the agent's
   own guard, and it is worked only in an attended session, so a human approves every change to it.
 - **It reverses no settled decision** recorded in an issue, a handoff, or the rule text.

.github/skills/backlog-burndown/SKILL.md

@@ -130,6 +130,13 @@ stands rather than being assumed cleared. The round's report names every issue i
 and the blocker each one waits on, since the maintainer reads the report rather than the issue
 bodies, and a stuck issue nobody names reads as ordinary backlog that simply never moves.
 
+An issue that needs a platform, by the test `session-handoff` "The Chain" states, is counted and
+is not ranked. It belongs to that platform's lane, so this run gives it no group, no worker, and no
+claim. An issue only part of which needs a platform is held back the same way. It waits on the
+split "The Chain" describes, which this run does not make, and the narrowed issue is ranked once
+that split lands. The round's report names each issue held back this way, its platform, and
+whether it waits on a split, as it names a `blocked` one.
+
 An issue that asks a question rather than states a defect is not ranked and is never guessed at.
 It has no group, no worker, and no claim, so nothing in "Raising a Blocked Question" applies to it
 except how the question travels. It goes to the maintainer at the end of ranking, per

.github/skills/session-handoff/SKILL.md

@@ -62,6 +62,26 @@ without trusting the parts of it that have gone stale.
 - **One open handoff issue per track per repository.** A track is a short kebab-case slug naming a
   lane of work, and a session that names no track uses `default`. Parallel lanes each name their
   own, so each closes its own predecessor and neither reads the other's state as current.
+- **A platform lane is for work that needs its platform.** A platform track is one kept for a single
+  host platform, such as `wsl` or `windows`. It takes an issue only when its defect reproduces, or
+  its fix can be verified, on that platform alone. It also needs that no Linux host outside WSL can
+  both fix the issue and verify the fix. An issue meeting both goes to its platform's lane. Any
+  other issue goes to the ordinary backlog, since a lane costs a separate chain and a host-bound
+  session. A defect in WSL interop, the `WSLInterop` binfmt handler, `/mnt/<drive>` paths, or
+  `wslpath` reproduces only inside WSL. So does running a Windows `.exe` from Linux, although WSL
+  runs on Windows. Behavior that differs under Windows PowerShell 5.1 or `winget`, or in Windows
+  path and ACL semantics, reproduces only on Windows. Either still goes to the backlog where a Linux
+  host outside WSL verifies its fix, as a plain unit test of a path translation can. An issue in
+  systemd user scopes, apt, a bash script, or a `pwsh` script run in a Linux container gets no lane
+  by its topic. The two conditions decide it like any other issue. Where only part of an issue needs
+  a platform, the session finding it files each such part as its own issue for its platform's lane.
+  The original is narrowed to the rest, or closed in favor of those issues where nothing is left. A
+  session that may not write issues names the unsplit issue in its report or handoff instead. An
+  issue for a platform whose lane was closed out reopens that lane with `new`, which chains onto the
+  closed link. A platform lane is closed out with no successor once none of its next steps needs its
+  platform. Before that, each remaining step with no issue of its own is filed as one. The lane's
+  last link carries the outcome as a comment, as `AGENTS.md` "Session Scope" closes a track whose
+  work is complete.
 - **The title is for humans**, shaped `Session Handoff [<track>]: <subject>`, which `new` composes
   from the track and the subject it is given, so what a session writes is the subject alone.
   Nothing parses the title, so a maintainer is free to rename one.

.github/skills/unattended-handoff/SKILL.md

@@ -154,7 +154,10 @@ does not qualify, since skipping one costs nothing and a guess costs a revert an
   and no comment asks the maintainer something still unanswered.
 - **The fix stays inside this repository's tree.** It changes no repository setting, ruleset,
   visibility, secret, or release condition, and needs no credential, account, or host the session
-  lacks.
+  lacks. The issue needs no platform, in whole or in part, by the test `session-handoff` "The
+  Chain" states. An issue that wholly needs a platform belongs to that platform's lane whatever host
+  the session runs on. One that partly needs it waits on the split that section describes, a write
+  the picker does not make.
 - **The fix edits nothing under the hub's `host-setup/agent-safety/`.** That tree is the agent's
   own guard, and it is worked only in an attended session, so a human approves every change to it.
 - **It reverses no settled decision** recorded in an issue, a handoff, or the rule text.