[EPAC-1958]: add estimate field to bugfix SPEC schema

.factory/prompts/bugfix-intake.md

@@ -25,6 +25,7 @@ You are running bugfix intake for `RiddimSoftware/epac`. Your job is to turn an
 - observed behavior
 - expected behavior
 - reproduction steps
+- estimate
 - evidence plan
 - validation plan
 
@@ -48,6 +49,7 @@ python3 scripts/intake/bugfix_spec.py new \
   --expected "The card explains that epac currently shows past debates and archival data." \
   --step "Launch epac." \
   --step "Open the Home tab." \
+  --estimate 8 \
   --evidence "Before/after screenshot of the Home status card." \
   --validation "Reporter confirms the TestFlight build no longer implies live data."
 ```

.factory/schemas/bugfix-spec.schema.json

@@ -14,6 +14,7 @@
     "affectedSurface",
     "observedBehavior",
     "expectedBehavior",
+    "estimate",
     "reproductionSteps",
     "acceptanceCriteria",
     "evidencePlan",
@@ -30,6 +31,11 @@
     "affectedSurface": { "type": "string", "minLength": 1 },
     "observedBehavior": { "type": "string", "minLength": 20 },
     "expectedBehavior": { "type": "string", "minLength": 20 },
+    "estimate": {
+      "type": "integer",
+      "enum": [1, 2, 4, 8, 16, 32, 64],
+      "description": "SF effort estimate from the exponential ladder 1, 2, 4, 8, 16, 32, 64. Opus maps this to Linear complexity as 1->1, 2->2, 4->4, 8->8, 16->16, 32->24, 64->40."
+    },
     "reproductionSteps": {
       "type": "array",
       "minItems": 1,

.factory/templates/bugfix-SPEC.md

@@ -31,6 +31,11 @@ This bugfix SPEC turns a contributor or LLM-session report into a factory-ready
   When [evidence plan runs]
   Then [artifact proves the behavior]
 
+## Effort estimate
+- Estimate: [1, 2, 4, 8, 16, 32, 64]
+- Confidence: [low|medium|high]
+- Reasoning: [One-sentence rationale]
+
 ## Evidence Plan
 [Screenshots, videos, XCTest/SnapshotTesting, Maestro flow, logs, manifest, or TestFlight validation needed to prove the fix.]
 

scripts/intake/bugfix_spec.py

@@ -17,6 +17,8 @@
 ROOT = Path(__file__).resolve().parents[2]
 DEFAULT_OUT_DIR = ROOT / ".factory" / "intake"
 TARGET_REPO = "RiddimSoftware/epac"
+EFFORT_ESTIMATE_LADDER = (1, 2, 4, 8, 16, 32, 64)
+EFFORT_ESTIMATE_CHOICES = ", ".join(str(v) for v in EFFORT_ESTIMATE_LADDER)
 
 REQUIRED_SECTIONS = [
     "## Original Prompt",
@@ -25,6 +27,7 @@
     "### Expected Behavior",
     "## Reproduction Steps",
     "## Acceptance Criteria",
+    "## Effort estimate",
     "## Evidence Plan",
     "## Validation Plan",
     "## Non-goals",
@@ -55,6 +58,7 @@ class BugfixSpecInput:
     steps: list[str]
     evidence: str
     validation: str
+    estimate: int
     original_prompt: str
     non_goals: str
     trace_id: str
@@ -93,6 +97,16 @@ def collect_steps(args: argparse.Namespace) -> list[str]:
     return collected
 
 
+def parse_estimate(value: str) -> int:
+    try:
+        estimate = int(value)
+    except ValueError as error:
+        raise ValueError(f"--estimate must be one of {EFFORT_ESTIMATE_CHOICES}") from error
+    if estimate not in EFFORT_ESTIMATE_LADDER:
+        raise ValueError(f"--estimate must be one of {EFFORT_ESTIMATE_CHOICES}")
+    return estimate
+
+
 def value_or_prompt(args: argparse.Namespace, attr: str, label: str) -> str:
     value = getattr(args, attr)
     if value:
@@ -111,6 +125,7 @@ def build_input(args: argparse.Namespace) -> BugfixSpecInput:
     expected = value_or_prompt(args, "expected", "Expected behavior")
     evidence = value_or_prompt(args, "evidence", "Required evidence")
     validation = value_or_prompt(args, "validation", "Reporter/TestFlight validation")
+    estimate = parse_estimate(value_or_prompt(args, "estimate", f"Effort estimate ({EFFORT_ESTIMATE_CHOICES})"))
     steps = collect_steps(args)
 
     missing = [
@@ -124,6 +139,7 @@ def build_input(args: argparse.Namespace) -> BugfixSpecInput:
             ("--expected", expected),
             ("--evidence", evidence),
             ("--validation", validation),
+            ("--estimate", str(estimate)),
         ]
         if not value
     ]
@@ -152,6 +168,7 @@ def build_input(args: argparse.Namespace) -> BugfixSpecInput:
         expected=expected,
         steps=steps,
         evidence=evidence,
+        estimate=estimate,
         validation=validation,
         original_prompt=original_prompt,
         non_goals=non_goals,
@@ -200,6 +217,11 @@ def render_spec(spec: BugfixSpecInput) -> str:
   When the reporter validates the affected surface
   Then their approval or blocker is recorded before release promotion.
 
+## Effort estimate
+- Estimate: {spec.estimate}
+- Confidence: medium
+- Reasoning: Initial SF effort estimate captured during intake; Opus may revise this during enrichment.
+
 ## Evidence Plan
 {spec.evidence}
 
@@ -295,6 +317,11 @@ def section_body(text: str, heading: str) -> str:
     return match.group("body").strip() if match else ""
 
 
+def section_field(section: str, label: str) -> str:
+    match = re.search(rf"^\s*-?\s*{re.escape(label)}:\s*(.+)$", section, re.MULTILINE)
+    return match.group(1).strip() if match else ""
+
+
 def validate_text(text: str) -> list[str]:
     errors: list[str] = []
     if not text.startswith("# Bugfix SPEC: "):
@@ -316,6 +343,16 @@ def validate_text(text: str) -> list[str]:
     if len(criteria) < 2:
         errors.append("missing at least two acceptance criteria starting with '- Given'")
 
+    effort_section = section_body(text, "## Effort estimate")
+    estimate_text = section_field(effort_section, "Estimate")
+    if not estimate_text:
+        errors.append("missing required effort estimate field: Estimate")
+    else:
+        try:
+            parse_estimate(estimate_text)
+        except ValueError as error:
+            errors.append(f"invalid effort estimate: {error}")
+
     if PLACEHOLDER_RE.search(text):
         errors.append("SPEC contains placeholder text; replace TBD/TODO/FIXME/bracket placeholders before intake")
 
@@ -409,6 +446,7 @@ def build_parser() -> argparse.ArgumentParser:
     new.add_argument("--surface")
     new.add_argument("--observed")
     new.add_argument("--expected")
+    new.add_argument("--estimate")
     new.add_argument("--step", action="append", default=[])
     new.add_argument("--evidence")
     new.add_argument("--validation")