MatchProof — Application Readiness

MatchProof is your application-readiness workspace. For each job you're targeting, it checks your resume against ATS parsing risks and against the job's requirements, one requirement at a time — and grounds every finding in your own evidence.

MatchProof measures how ready your application is, not how likely you are to get the job. It is preparation support, never a hire probability or an ATS pass guarantee.

Target applications

The MatchProof home lists your target applications — "each target application is one job you're applying to." Its resume, job description, and analysis all live inside it.

• New target application — give it a title (required), plus an optional company and seniority. Your resume and job description stay in the ReadyLoop workspace the application is built on; MatchProof analyzes them together without a second copy.
• Every target application is backed by a job in the Jobs section, so the Jobs list always shows every job you're pursuing in one place.
• View on a target card opens the target's Application Cockpit directly — no redirects, the application opens with its real id and its own data.
• Each row shows an Active / Archived badge and a Resume attached / No resume yet badge.
• Edit — update the title, company, or seniority. If someone changed it elsewhere, you'll be asked to close and reopen it before trying again ("This target application was changed elsewhere — close and reopen it, then try again.").
• Duplicate — copy an application, handy for similar roles.
• Archive / Restore — archiving tucks an application away but keeps it recoverable.
• Delete — permanent ("Delete this target application?" → Delete permanently). Archive instead if you might want it back.

Two entry cards complete the home: Career Evidence Vault and Production certification, each with an Open badge.

Application Cockpit

Open a target application to land on its Application Cockpit — the home for that application's readiness. The cockpit is built on one authoritative status summary that spans MatchProof and ReadyLoop, so you always see the true state of the whole application package:

• The status strip — four chips: Application, Evidence, Blueprint, and Certification, each with its real status ("never computed", "in progress", "partial", "complete"). The chip vocabulary is the system's own — nothing is claimed until it's actually done.
• Issue count — how many readability issues your latest analysis found (shown only when there are any).
• Prep plan line — when a ReadyLoop prep plan is active, the interview date shows right here.
• Manage resume & JD in ReadyLoop — jumps to the workspace that owns the resume and job description.
• Next Best Actions — the ranked cross-product action catalogue for this target (categories: Application, Evidence, Interview, Practice, Commercial), with an Open button.
• One ranked next step — the "Next best step" card names the single most important thing to do next, with a CTA like Improve Application (or Re-run analysis when nothing is pending). The priority comes from the backend's deterministic ranking.
• Application Readiness cards for each analyzed dimension, with per-dimension badges ("Clean" / "Needs attention" / "Blocked" / "Not in this analysis" for readability; "Analyzed" / "No requirements yet" for job match; and so on), each with its drill-down link (Review findings, Review requirements, Review evidence, Review recruiter view, Open Resume Studio, Review ranked actions, Review claim exposure, Review defensibility). A Top actions list follows.
• Re-run analysis — each run writes a new assessment pinned to the current resume and job description; earlier results are never edited. The re-run strip says so: "Re-run writes a new assessment pinned to the current resume and job description — earlier results are never edited."
• Dimensions that aren't built yet appear honestly under "Coming in later sprints" with a Planned badge and the sprint they're planned for.

On a fresh application, the empty card says "No analysis yet" with a Run analysis button.

The readiness sections

• ATS Readability — how reliably your resume can be machine-read, checked on the latest uploaded version. Status is Clean, Needs attention, or Blocked; blocking issues can stop an ATS from parsing the resume at all. Each finding shows why it was flagged, its fix, and a "Blocking"/"Worth checking" badge. The page is explicit that this is "not a verdict on you or your application". Analyses keep a history you can archive or delete.
• Semantic Job Match — each job requirement matched against evidence already in your resume, gaps and strengths side by side. Must-haves and preferred requirements are counted as matched, weak, or missing; each is marked Matched / Transferable / Weak / Missing. Requirements you marked not relevant are excluded, and the page says so.
• Evidence Strength — how strongly each requirement is proven by what your resume already says (strong / adequate / thin / missing), rated across dimensions like ownership, specificity, scale, measurable outcome, relevance, recency, and impact. The Truth Guard never guesses: "Missing evidence — check with you, never guess." If your resume doesn't prove something, it asks — Yes, I have it / No / Not sure — and records what you confirm (category, what you did and with what result, when). Confirmed evidence can be changed or removed; genuine gaps are pointed at ReadyLoop to prepare.
• Recruiter Impact — what a recruiter can see in the first 20 seconds: your strongest resume-backed signals and the likely concerns (with concern badges like "You noted a gap" / "Not obvious from resume" / "Loosely evidenced"), each with what would strengthen it.
• Improvement Actions — the smallest set of changes that most increase the application's quality, ranked High / Medium / Low priority with an honest status vocabulary: Open / Accepted / Completed / Dismissed / Superseded. Each action explains why it matters and how to do it. Regenerate rebuilds them; when nothing valuable exists, the engine says so.
• Claim Exposure — the claims in your resume most likely to attract interview scrutiny, extracted verbatim and ranked High / Medium / Low exposure, each with its requirement overlap, resume line, and the "Why this exposure?" breakdown — including whether saved evidence covers it yet.
• Claim Defensibility — the likely follow-up probes for each of your claims, checked against the evidence you have saved ("{covered}/{total} probes covered"), with what to prepare and the next step. Practice this claim opens Claim Practice.
• Claim Practice — voice rehearsals of your highest-exposure claims: the opener is the claim itself, then every predicted follow-up. Sessions carry honest statuses (Preparing / Listening / Thinking / Paused / Reconnecting / Completed / Cancelled), difficulty badges, and an Invalidated state when the claim was re-analyzed after the practice started.
• Rewrite Suggestions — wording and structure improvements assembled only from your resume, the job requirements, and your confirmed evidence (kinds: Concise rewrite, Strengthen, Bullet restructure, Section order, Needs your input). Each shows Current vs Suggested and its evidence source; Apply to resume, Write and apply, or Reject. Suggestions are deterministic — the same resume, requirements, and evidence always produce the same suggestions.
• Resume Studio — edit, compare, undo, and export a job-specific resume with version history (Original / AI rewrite / Manual edit, with Set active and Undo; versions are immutable). Commit as a new version pins your draft; Export builds DOCX/PDF from a template with Download / Preview / Regenerate.
• Reuse career evidence — apply confirmed vault facts into an application ("Apply selected ({n})"); nothing unconfirmed or rejected is offered, and already-linked facts are marked.

Career Evidence Vault

Your confirmed experience, skills, and stories, organized into a career graph that compounds across every target application. Nothing is treated as fact until you confirm it.

• Build my career graph to pull in your confirmed evidence and stories from across applications. Nodes are grouped by kind: organizations, roles, projects, achievements, measurable outcomes, skills, stories, and responsibilities.
• Node badges are honest: Confirmed, Corrected by you, Rejected, or Unconfirmed — AI-derived. Confirm, correct, reject, split, or delete each one; your corrections always win over re-ingests.
• Possible duplicates can be merged into one node, and a node can be split into a new one you own.
• Audit history records every confirm, correct, reject, delete, merge, and split — deleted content never resurfaces.
• Refresh from evidence re-ingests your confirmed evidence at any time.

Production certification

The certification page summarizes preparation completeness across your applications and career vault. Compute certification runs the checks (the button reads Recompute once a snapshot exists); you'll see "Ready for release prep" or "Not fully ready yet" with an "All checks passed" / "{n} open" badge, the list of checks ("Every check must pass for overall readiness."), and an immutable history — "Immutable snapshots, newest first." Again: this measures preparation completeness only — not a probability of hire, and not an ATS acceptance guarantee.

Getting started

1. Open MatchProof and create your first target application.
2. Make sure its ReadyLoop workspace has your resume and the job description.
3. Open the application and Run analysis.
4. Follow the Next best step and Top actions into the section pages.
5. Build your Career Evidence Vault so every application gets stronger evidence over time.

See also: Career Vault — the source of the confirmed evidence MatchProof works from, and ReadyLoop — where each target application's resume and job description live.