科研 Skill

项目范围核查

scope

Interrogates assumed project scope before a civil-engineering proposal is drafted, surfacing the real problem and scope options. Use before quoting, proposing, or committing to a scope of work.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Scope Check

Pre-proposal interrogation protocol for [Your Organisation]. Forces the real problem to the surface before scope, fee, or methodology is committed.

Adapted from the gstack /office-hours methodology for the firm engineering and inspection practice.

RULE: No proposal draft, no fee estimate, no methodology selection — until this interrogation is complete. Scoping before framing is how the firm ends up underpaid for overdelivery.


Phase 1 — Receive and Restate

When the user describes a potential project or client request:

  1. Restate what you heard in one sentence — not what was said, but what was meant: the problem the client has, not the service they asked for.
  2. Flag the gap between what the client asked for and what they actually need (if one exists).

Example:

Client asked for: "Tank calibration for two tanks" What they actually need: "A defensible volumetric baseline for fiscal custody transfer compliance — calibration is the method, not the goal"

This reframe may change the scope, the deliverable, and the fee.


Phase 2 — Six Forcing Questions

Work through these in order. Ask only the ones that are unanswered from the context already provided. Do not ask questions the user has already answered.

Q1 — What is the actual consequence if this is not done?

Not the project rationale — the real downstream consequence for the client.

  • Is this regulatory / compliance-driven?
  • Is it insurance / underwriting-driven?
  • Is it a precondition for another project?
  • Is the client under pressure from a principal (operator, JV partner, bank)?

This determines urgency, leverage, and whether the firm is solving a pain or a preference.

Q2 — What does "success" look like to the client — in their words?

Not what the deliverable is, but what the client will say when the job is done well.

  • A signed-off report? A passed inspection? An insurer satisfied? A claim dismissed? A plant back online?
  • If you don't know this, you cannot scope the right deliverable.

Q3 — What has already been tried or done?

  • Prior inspections, assessments, repairs?
  • Other firms engaged previously?
  • What did they miss, or why was the client unsatisfied?
  • Are there existing drawings, reports, or data available?

This prevents the firm from re-doing work and charging for it, and may reveal why the client is coming to the firm specifically.

Q4 — What is the actual constraint — time, access, or budget?

Pick the binding one.

  • If time is the constraint: methodology must be fast; deliverable may be preliminary.
  • If access is the constraint: methodology must fit the access window; mobilisation timing matters more than fee.
  • If budget is the constraint: scope must be reduced to fit, or client must be educated on minimum viable scope for defensible output.

Do not guess. Ask directly if not already known.

Q5 — Who else is quoting, and on what basis?

  • Is this a competitive tender or a direct award?
  • Are competitors quoting the same scope, or a different interpretation?
  • Is there a BOQ / scope document, or is each firm defining its own scope?

This determines whether to price to win, price for margin, or decline.

Q6 — What is the riskiest assumption in the current scope framing?

The thing that — if wrong — would make the proposal embarrassing or loss-making.

  • Assumed access that may not exist?
  • Assumed data that the client may not have?
  • Assumed structural condition that may require additional investigation?
  • Assumed regulatory pathway that may not apply?

Name it explicitly before scoping.


Phase 3 — Reframe and Alternatives

After the six questions, generate:

3.1 Reframed Problem Statement

One paragraph: what is the firm actually being asked to solve, in engineering terms, not in service-catalogue terms.

3.2 Three Scope Options

Present three interpretations of what the firm could deliver:

OptionScopeFee SignalRisk
NarrowMinimum defensible deliverable that answers the client's core questionLowestScope creep risk if client expects more
StandardFull scope as originally framedMidCurrent baseline
ExpandedBroader scope that solves the underlying problem, not just the stated requestHigherClient may not have budget; but this is the right answer

Recommend one. Justify it in one sentence.

3.3 Scope Exclusions (Draft)

List 3–5 items that are explicitly out of scope under the recommended option. These go into the proposal to prevent scope creep.

3.4 Key Assumption Register (Draft)

List 3–5 assumptions the firm is relying on. Each becomes a caveat in the proposal.


Phase 4 — Methodology Check

Once scope is agreed, challenge the methodology:

  1. Is the proposed method the most defensible for the client's purpose (fiscal, regulatory, insurance, litigation)?
  2. Is it the most efficient given the access and timeline constraints?
  3. Is there a better method the firm could offer that competitors may not (SHM, digital twin, integrated NDT + structural analysis)?
  4. Does the method produce output in the client's required format (API standard, [National Regulator] format, operator template)?

Flag any mismatch between method and purpose before the proposal is drafted.


Output — Scope Brief

At the end of this interrogation, produce a one-page scope brief:

PM SCOPE BRIEF
Ref: PRJ/SB/[YYYY]/[NNN]
Date: [date]
Client: [name]
Project: [working title]

REAL PROBLEM:
[one paragraph reframe]

RECOMMENDED SCOPE OPTION: [Narrow / Standard / Expanded]
RATIONALE: [one sentence]

KEY SCOPE EXCLUSIONS:
1.
2.
3.

KEY ASSUMPTIONS:
1.
2.
3.

RISKIEST ASSUMPTION: [from Q6]

BINDING CONSTRAINT: [time / access / budget]

RECOMMENDED NEXT STEP: [draft proposal / request more info / decline / site visit first]

Prepared by: [PM]

This brief feeds directly into proposal drafting (via the project's document-drafting tool or template). Do not begin the proposal until this brief is reviewed and confirmed.

示例提示词

客户想加一项检测服务,帮我先核查清楚这算不算在范围内。

所属插件

建筑学与土木工程面向建筑与土木研究和设计:场地分析与方案设计、Revit/AutoCAD/Rhino-Grasshopper 与 OpenBIM-IFC、机电与结构协调、消防与无障碍、围护结构与日照能耗、造价与工程量,以及土木工程项目的进度、质量、安全、风险与 RFI 管理。

安装该插件会一并获得同领域的 44 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

项目范围核查能做什么?

Interrogates assumed project scope before a civil-engineering proposal is drafted, surfacing the real problem and scope options. Use before quoting, proposing, or committing to a scope of work。

如何安装项目范围核查?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「项目范围核查」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

项目范围核查需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

项目范围核查经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。