executive-application-helper

Run a three-axis integrity check on the framework's evidence-bank files at the end of any session that touched the bank — Examples_Master.md, the Section 4/5/6/7 files, any Examples_Section_A/B/C/D companion file — plus the append-only files that actually truncate in practice: both changelogs (evidence-bank changelog + current weekly archive, instructions changelog) and the Project_Instructions file. Use at post-draft maintenance (STEP 4) before the maintenance report, OR when the candidate says 'check the bank', 'integrity check', or 'make sure I didn't break anything'. Mechanical (~30–60 sec via bash + grep); catches three failure modes that otherwise go undetected for weeks: end-of-file truncation, cross-reference breakage, and silent line-count drops.

Apply the candidate's executive voice and style — written or spoken — or audit existing material against it. Use this skill any time you are (a) drafting, rewriting, sharpening or reviewing prose the candidate will send or sign (emails, cover letters, statements of claims, briefings); (b) preparing for or rehearsing an interview the candidate will sit (recruiter screens, behavioural panels, executive panels — including STAR-style prep documents and opening scripts); or (c) generating an interview-prep document using the framework's eight-section template. Also use when the candidate asks to 'make this sharper / more like me / in my voice' or for a critique, audit or tone review of any writing. Default to invoking this skill rather than freelancing on voice, tone or interview-prep structure.

Verify that a panel brief for an SES or equivalent senior role contains a complete capability mapping section before delivery — all framework capability groups mapped per question, Strong/Adequate/Watch ratings on every cell, a named pivot story for every Watch, an overall coverage matrix, a values/behaviours cross-frame, and the cumulative-impression reminder. Use this skill whenever a panel or executive-panel brief for an APS SES role (or a state-government / board-level equivalent scored against a formal capability framework such as the ILS) is about to be saved to Interview Prep/ or delivered. Also use when [Candidate First Name] asks 'check the capability mapping', 'is the ILS coverage complete', or 'run the mapping check'. The check is mechanical (~30–60 sec) and catches the panel-prep failure mode where coverage thins on one capability and no pivot story exists when the panel probes it.

Generate the STEP 2 coverage analysis for a drafted application document — mapping the draft against every section of the source materials (candidate information pack, position description, advertisement, selection criteria, recruiter requests, 'About You' / 'Key Accountabilities' / 'Capabilities' sections) and marking each Addressed / Partially Addressed / Not Addressed with the supporting evidence-bank entry IDs. Use this skill immediately after any application draft is finalised (cover letter, statement of claims, targeted pitch, selection criteria response, recruiter response) — STEP 2 is mandatory for every application-drafting session. Also use when [Candidate First Name] asks 'how well does this cover the PD', 'coverage analysis', 'what did we miss', or 'map the draft to the criteria'. Produces the Step 2 deliverable tables directly.

Verify that a drafted application document complies with the framework's word-count and character-count discipline before delivery — document-type word bands, any matrix-row or brief-stated override, and (critically) per-question hard character limits on job-platform screening questions (e.g. Seek's 1,000 characters per question, which the platform truncates silently). Use whenever a draft is about to be saved to Submissions/ or presented as final output. Also use when the candidate asks to 'check the length', 'count words', or 'is this within the limit'. Mechanical (~5–10 sec per document).

Confirm that a drafted application document actually follows the prescription of its matched evidence-bank matrix row — every B-ID in the row's body stack is reflected in the draft, the dominant A-ID positioning leads, the template and length band match, and the fit-acknowledgment style and AI register are honoured. Use this skill at Step 1 close on any application-drafting session where a matrix row was matched (cite by ID), alongside length-check and watchouts-sweep. Also use when [Candidate First Name] asks 'did we follow the row', 'trace the draft to the matrix', or 'is this on-prescription'. Catches prescription drift: a draft that silently swapped the row's tested body stack for ad-hoc evidence selection, losing the calibration the row encodes.

Run an honest timing audit on the spoken responses in the candidate's interview-prep documents, at the candidate's actual measured speaking pace — never at the generic 165 wpm executive baseline. Use this skill whenever a speaking-notes addendum or prep document with spoken responses is about to be delivered, AND after every content reshape or tightening pass (timing drifts with every edit). Also use when [Candidate First Name] asks 'how long does this run', 'time these responses', 'am I over budget', or 'run the pace check'. The audit is mechanical (~10–30 sec): word count ÷ actual wpm vs the format's time budget, with a two-pass tightening plan for any response that runs over.

Verify at session close that any sensitive personal context [Candidate Name] shared during the session (health, medical history, family circumstances, or anything on the sensitive-personal-information list) was used ONLY in the named deliverable it was shared for, with his explicit consent, and was NOT propagated to the evidence bank, the memory system, or staged for any future deliverable. Use this skill at the close of any session in which sensitive personal context surfaced — interview prep with lived-experience questions is the canonical case. Also use when [Candidate First Name] asks 'check the discretion rule', 'did anything personal leak', or 'verify the personal-context boundary'. This is a blocking check: a propagation finding must be reversed before session close, not noted for later.

Verify — BEFORE any structural edit to the candidate's evidence-bank, project-instructions or changelog files — that a Convention 13 pre-edit snapshot exists in _resources/ under the dated naming convention, and that the _resources/ retention discipline (keep 3 most-recent good backups, _archive/ for older, _quarantine/ for truncated, RESTORE_INDEX.md current) holds. Use this skill at the START of any session step that will add or move a matrix row, change entry format, restructure a Section file or Examples_Master.md, restructure Project_Instructions_V*.md or either changelog, or run a multi-edit Python script across the bank. Also use when [Candidate First Name] says 'check the snapshots', 'is there a backup', or 'verify the restore index'. This is a blocking gate, not advisory: a structural edit without a pre-edit snapshot is unrecoverable if it truncates.

Audit STAR responses in the candidate's interview-prep documents against his STAR construction conventions before delivery. Use this skill whenever an interview-prep document (recruiter screen, behavioural panel, executive panel) containing STAR responses or a speaking-notes addendum is about to be saved to Interview Prep/ or presented as the final output of a prep session. Also use when [Candidate First Name] says 'audit the STARs', 'check the STAR structure', 'are these STARs right', or 'run the STAR check'. The audit is mechanical (~30–60 sec) and catches the most common prep failure modes: an Action written as a what-list instead of a how-mechanism, numbered-list structure instead of flowing 5.8-style prose, missing or unbolded principle lines, and spoken responses without a drop list of at least 20 seconds of cuttable runway.

Cross-check a drafted application document (cover letter, statement of claims, targeted pitch, recruiter response, screening-question response) against the candidate's watch-outs register (Examples_Section_7_Watchouts.md) before delivery. Use whenever a draft is about to be saved to Submissions/ or presented as the final output of an application-drafting session. Also use when the candidate asks to 'sweep this for watchouts', 'check this for overclaim', or 'verify this against Section 7'. The sweep is mechanical (~30–60 sec) and catches the most common executive-application failure mode: an overclaim that the register would have caught.