fix(#852): add consumer-completeness check to code-review Correctness dimension

Add a "Consumer completeness" bullet to the Correctness section of the
code-review skill. When a PR adds new values to an enum, dispatch table,
JSON schema enum, or case/switch structure, the review agent will now
check that all downstream consumers handle the new value — including
scripts, configs, and files not in the diff. This closes the gap that
allowed PR #765 to be approved despite a missing handler for the new
`reject` disposition in `post-review.sh`.

Note: `make lint` could not run due to sandbox Go toolchain permission
error (unrelated to this change). The change is markdown-only.

Closes #852

Signed-off-by: fullsend-code <fullsend-code@users.noreply.github.com>

Author: fullsend-code

internal/scaffold/fullsend-repo/skills/code-review/SKILL.md

@@ -66,6 +66,11 @@ dimension carry over to another — each requires its own scrutiny.
   external API (GitHub, cloud providers, etc.), verify the API accepts
   the new values for every code path that calls the function. Different
   API operations often have different required fields.
+- Consumer completeness: if the change adds new values to an enum,
+  dispatch table, JSON schema enum, or case/switch structure, identify
+  all code paths that consume or branch on that type (including scripts,
+  configs, and files not in the diff) and verify each handles the new
+  value. A new variant with no downstream handler is a logic error.
 - Test adequacy: are the right behaviors tested?
 - Do the tests actually constrain the code's behavior, or do they
   merely assert it runs?