Specwarden lockupspecwarden

The deepest moat in our reviewer

Detect the Tier-2 supplier-blame pattern auditors hate — automatically

When a DFMEA cause reads "cell supplier delivers out-of-spec parts," that's not a design-side cause. It's the most-flagged audit failure in AIAG-VDA reviews. In testing, Specwarden's AI caught this pattern 6 out of 6 times — even on rows where we didn't plant it.

How it works

Upload your DFMEA

Drop in your spreadsheet — .xlsx, .xls, .csv, or .docx. Specwarden detects AIAG-RPN or AIAG-VDA format automatically. No configuration.

Review runs in under 60 seconds

7 deterministic rule checks run first, then Claude Sonnet reads every row for supplier-blame causes, vague language, and missing rationale.

Get a report you can act on

Download a PDF or Word report with every finding ranked by severity and a specific fix recommendation. Ready to attach to your design package.

Why Specwarden catches supplier-blame

100%

precision on spontaneous catches

In our 10-fixture eval suite, Specwarden caught Tier-2 supplier-blame on 3 rows we hadn't planted. A senior manufacturing QE rated all 3 catches as correct — zero false positives.

D-001b

a dedicated moat check ID

Supplier-blame detection is not a generic "vague text" check — it's a dedicated rule grounded in AIAG-VDA Handbook §1.4.1 and §2.4, trained on the semantic distinction between design-spec and supplier-execution causes.

Free

on your first 30-row DFMEA

See the supplier-blame catches on your own data. Free tier includes 5 visible findings — if Specwarden catches a D-001b, you'll see it without paying.

How the moat detection works

Tier-1 PFMEA contamination (D-001a)

When a DFMEA row's Failure Mode describes an assembly-line operator error or a manufacturing process step — that's PFMEA scope. Specwarden flags it and suggests the correct reframe.

Tier-2 supplier-blame reframe (D-001b)

A Failure Cause that blames the upstream supplier for delivering bad parts isn't a design cause — it's a procurement and incoming inspection problem. Specwarden identifies the design-specification gap that the cause should be pointing at instead.

Recommended Action reframing

Supplier-blame causes usually lead to weak Recommended Actions ("source from better supplier"). Specwarden flags the action alongside the cause and suggests the design-side corrective (e.g., "add minimum acceptance criterion to the engineering drawing").

Built for engineering teams reviewing high-stakes technical documents.

Workflow proof

Reviews engineering documents against risk and compliance criteria

Use-case proof

Designed for FMEA, specs, quality checks, and design review workflows

Beta proof

Private beta feedback from technical reviewers and operators

Empirical result — Specwarden internal testing

Across 10 hand-authored DFMEA test cases in our evaluation suite, Specwarden spontaneously caught Tier-2 supplier-blame reframes on 3 cases — on rows where we had not specifically planted the pattern. Our internal reviewer (a senior manufacturing quality engineer) rated all 3 catches as correct: 100% precision on spontaneous detection.

The DFMEA test cases included foundry casting preheat drift, NADCAP procurement actions, and casting porosity — none were authored as "supplier-blame tests." Specwarden found them anyway.

Read the full technical breakdown in our article on Tier-2 supplier-blame in DFMEA.

Questions about supplier-blame detection

What does "Tier-2 supplier-blame" mean exactly?
In a DFMEA, every Failure Cause should describe a gap in the design specification — something the design engineer could fix by changing the drawing or tolerances. A "Tier-2 supplier-blame" cause blames the upstream supplier's execution instead. AIAG explicitly classifies this as an audit failure because the design team can't act on it.
Why is this pattern so common?
Two reasons: first, it's genuinely easier to write "supplier delivered bad parts" than to diagnose the design specification gap that permitted those parts. Second, many DFMEAs are written by engineers who've never been through an AIAG-VDA audit and don't know the pattern is a red flag.
How is Specwarden different from a generic AI tool?
Generic AI tools detect obvious placeholder text ("TBD", "N/A") but don't understand the semantic distinction between a design-specification cause and a supplier-execution cause. Specwarden's reviewer is grounded in AIAG-VDA Handbook §1.4.1 and §2.4 — it knows what a DFMEA cause is supposed to say.

See the supplier-blame catches on your DFMEA.

Free. 30 rows. No card. Takes 60 seconds.

Join waitlist