forgeplan_estimate

Estimate effort (in hours) for an artifact by parsing its Functional Requirements and Implementation Phase checkboxes, then applying a multi-grade profile (Junior / Middle / Senior / Principal / AI). Returns a confidence-scored breakdown per grade, optionally using LLM-based complexity scoring or manual overrides.

Category: Quality

When an agent calls it

Sprint planning — get hours for each PRD / RFC before committing scope.
Grade matching — see how many hours the task takes for your own grade via my_grade.
AI-assisted plans — grade: "ai" applies AI multipliers (typically 0.03–0.4× of senior) for tasks that can be delegated.
Calibration feedback loop — compare actual vs estimated after the sprint to tune the profile.

Input parameters

Name	Type	Required	Description
`id`	`string`	yes	Artifact ID to estimate.
`grade`	`string`	no	Override grade for all items: `junior`, `middle`, `senior`, `principal`, `ai`.
`my_grade`	`bool`	no	Auto-detect grade from config `grade_profile` + artifact domain inference.
`llm_score`	`bool`	no	Use LLM-based complexity scoring instead of rule-based heuristics.
`complexity`	`string`	no	Manual complexity overrides, e.g. `"FR-001=5,FR-002=3"`.

Schema source: crates/forgeplan-mcp/src/server.rs::EstimateParams

Returns

{
  "id": "PRD-039",
  "total_items": 12,
  "by_grade": {
    "junior": { "hours": 96, "confidence": 0.6 },
    "middle": { "hours": 52, "confidence": 0.75 },
    "senior": { "hours": 28, "confidence": 0.85 },
    "principal": { "hours": 18, "confidence": 0.8 },
    "ai": { "hours": 4.2, "confidence": 0.5 }
  },
  "breakdown": [
    { "item": "FR-001", "complexity": 3, "senior_hours": 4 }
  ]
}

Example invocation

{ "id": "PRD-039", "my_grade": true }

Typical sequence

forgeplan_estimate with my_grade: true — get personal hours.
Compare with sprint capacity — scope down if needed.
After sprint: feed actual vs estimate back into grade_profile config.

CLI equivalent

forgeplan estimate PRD-039 --my-grade
forgeplan estimate PRD-039 --grade ai