skills: cut blabla and redundancy from skill prose

Per-skill prose tightening from the same audit swarm; each finding
passed an adversarial second reviewer. No behavioural change.

- audit: drop garbled "Conventions require deferred audits to
  compound" sentence (says the opposite of intent; the preceding
  line already closes the loophole).
- brainstorm: drop forward-pointing meta-comment about the
  Rationalisations table.
- debug: drop third restatement that debugger.md is the single
  source for the carrier/handoff fields.
- docwriter: drop motivational opener; Overview starts at the
  waste argument.
- fieldtest: fold the 2-4-examples rationale into the dispatch
  sentence instead of restating the count a fourth time.
- implement: drop "known platform constraint at the time" aside.
- planner: cut the verbose anti-drift paragraph (which restated
  the very table it claimed not to) down to a cross-reference.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

audit/SKILL.md

@@ -25,9 +25,7 @@ explicit backlog issue (issue tracker configured under
 - the reason for deferral,
 - the date the audit will be re-run.
 
-"We want to keep moving" is not a valid reason. Conventions
+"We want to keep moving" is not a valid reason.
-require deferred audits to compound, which is the failure mode
-the skill prevents.
 
 Also triggered manually when baseline drift is suspected
 outside a cycle close (e.g. the bencher reports an unexplained

brainstorm/SKILL.md

@@ -42,10 +42,6 @@ Triggers:
 - A cycle described as "small, only two iterations".
 - A cycle where you are confident you know what to build.
 
-The first two cases above are the most common pressure to
-skip; both are explicitly resisted by this skill (see Common
-Rationalisations).
-
 ## The Hard-Gate
 
 ```

debug/SKILL.md

@@ -87,11 +87,8 @@ separate one.
 
 - **Agent dispatched:** `agents/debugger.md` — carries the
   four-phase process, the Phase 4.5 escalation rule, the
-  Common Rationalisations table, and the Red Flags list. It is
+  Common Rationalisations table, and the Red Flags list. The
-  also the single source for the carrier and handoff field
+  orchestrator does not execute these phases directly.
-  definitions (Carrier contract / Output format); this skill
-  names the fields but does not redefine them. The orchestrator
-  does not execute these phases directly.
 - **Hand-off target:** `../implement/SKILL.md` — runs the
   GREEN side after the orchestrator has decided whether to
   commit the RED test separately or as part of the final fix

docwriter/SKILL.md

@@ -9,10 +9,8 @@ description: Use when the API surface of one or more components has stabilized a
 
 ## Overview
 
-API documentation rots silently. Every iteration changes APIs
+Running this sweep per-cycle is waste — documenting an item
-and module boundaries; doc comments lag. Running this sweep
+that gets renamed two iterations later just burns context. The right moment is
-per-cycle is waste — documenting an item that gets renamed two
-iterations later just burns context. The right moment is
 post-stability: after a stretch of cycles in which the surface
 in question has held still. This skill is the third bucket in
 the cadence taxonomy (per-cycle-mandatory audit; orchestrator-

fieldtest/SKILL.md

@@ -115,17 +115,14 @@ carrier.
 
 Dispatch `fieldtester` with the carrier from the Handoff
 Contract below. The agent picks 2-4 examples (one per axis
-the cycle touched), implements them as downstream consumers,
+the cycle touched; fewer loses the signal on variation, more
-runs them, classifies findings, and writes the spec. All
+overflows one readable report), implements them as downstream
-artefacts (fixtures + spec) stay in the working tree as
+consumers, runs them, classifies findings, and writes the
-unstaged changes; the orchestrator commits them after
+spec. All artefacts (fixtures + spec) stay in the working tree
+as unstaged changes; the orchestrator commits them after
 reviewing the report (suggested commit subject:
 `fieldtest: <cycle> — <N> examples, <K> findings`).
 
-Two to four examples is the right size. One is too few (no
-signal on variation); five is too many for one report to
-stay readable.
-
 ## Handoff Contract
 
 `fieldtest` consumes (from orchestrator at cycle close) the carrier

implement/SKILL.md

@@ -243,8 +243,7 @@ independently at cycle close.
   does not permit a subagent to spawn other subagents. The
   orchestrator-agent cannot dispatch the role-agents above;
   it adopts each role as a sequential phase in its own
-  context. This was a known platform constraint at the time
+  context.
-  the role was designed.
 - **Input sources:**
   - `../planner/SKILL.md` — produces the plan files this
     skill consumes

planner/SKILL.md

@@ -72,13 +72,9 @@ Read the returned file-map. Lift the relevant entries into
 the plan's "Files this plan creates or modifies" section; the
 orchestrator may add or amend entries that recon missed.
 
-Dispatch carrier — the fields `spec_path`, `iteration_scope`, and
+Dispatch carrier: see the **Carrier contract** table in
-`focus_hint`, defined authoritatively under **Carrier contract**
+`agents/plan-recon.md` (single source for `spec_path`,
-in `agents/plan-recon.md` (which marks `spec_path`/`iteration_scope`
+`iteration_scope`, `focus_hint`).
-mandatory and `focus_hint` optional, and specifies the
-BLOCKED-on-missing rule). That table is the single source for those
-fields; it is deliberately not restated here, so the two files
-cannot drift.
 
 Recon does NOT decide decomposition. Two recon dispatches per
 planner run is unusual but legitimate (e.g. one for the main