em-bigmac-framework v0.10.1 Delivery

Qualitara's Big Mac best-practices framework, automated - run a full account-wide framework audit into one HTML grid (bigmac-run), fill the rest by hand in a friendly local form (soft-audit), and run sprint-ceremony audits with client, internal and board reports (sprint-audit).

Problem it solves: objective, audit-grade evaluation of a team's Agile ceremonies against a formal framework, producing tailored reports for the team, the client and Qualitara leadership - removing EM subjectivity.

soft-auditUser skillcode-backedLight · ~5-20k

The manual auditor. Generates a friendly, self-contained local HTML form where the EM scores every not-yet-automated test Pass / Fail / N/A with optional evidence - per-category filters, search, "Unanswered only", a progress bar, and answers that auto-save in the browser. Tests already covered by an automated auditor are pre-filled and locked. One-click Export produces an answers JSON that is processed into a method: manual results file the run merges, and the report regenerates.

Read the full guide & setup
Reference - what it brings, its process, and what it needs
BringsA fast, low-friction way to complete the framework where no automated auditor exists yet - so the whole-framework pass rate becomes real, while the dashboard still separates automated from manually-attested coverage.
Process
  1. Generate: render the form pre-filled with auto-scored tests (locked).
  2. Fill: the EM scores the rest in a browser; progress saves locally.
  3. Export: one-click answers JSON.
  4. Process: validate + write the manual results file (auto-covered ids skipped), then regenerate the report.
NeedsPython bigmac-runRead/Write/Basha browser to fill the formmanifest bigmac_run block
bigmac-runUser skillcode-backedMedium · ~20-80k

Audits a whole account against the complete Big Mac Framework v2 (233 atomic tests across Engineering, QA and Product Management) and renders one local HTML replica of the Airtable grid - every column, every test. It is the aggregator: it syncs the framework, collects results from capability-auditor skills, fills the grid, and marks every not-yet-automated test In progress. The report's "Automation coverage" section documents every capability auditor, live and planned.

Read the full guide & setup
When to use it
run the Big Mac Frameworkrun the framework auditgenerate the full framework reportrefresh the BigMac runsee overall framework coverage
How it works
  1. Resolve paths from the manifestReads the renderer, data, and report folders from the account's bigmac_run block.
  2. Sync the frameworkPulls the Framework v2 records from Airtable and normalizes them into the full test grid with all its metadata.
  3. Collect capability resultsFor each live auditor (today: sprint-audit), runs its adapter to read the latest evaluation and emit a results file.
  4. Render the gridFills the full test grid - real results for audited tests, "In progress" for every test not yet automated.
Example
You say
"run the Big Mac Framework for Acme Robotics" - the Scrum Management capability already has a recent sprint-audit run.
You get back
An HTML replica of the Airtable grid: pass/fail/N-A for the Scrum Management tests from the latest sprint-audit, "In progress" for every other capability, and a coverage section documenting which auditors ran.

Gotcha: It is read-only on the sprint-audit data - the run never writes back to the evaluation.

Gotcha: If sprint-audit has never run for the account, its tests render "In progress" like the rest.

Reference - what it brings, its process, and what it needs
BringsOne account-wide view of framework compliance and its automation roadmap in a single artifact - real Test Results where an auditor has run, "In progress" where one has not, and the owning skill named on every row.
Process
  1. Sync: pull Framework v2 from Airtable, normalize to a local file.
  2. Collect: run each live capability auditor's adapter (sprint-audit -> Scrum Management today, read-only).
  3. Assemble: fill every test from the results, default the rest to In progress.
  4. Render: the full grid + coverage roadmap to HTML.
NeedsAirtablePython bigmac-runRead/Write/Bashmanifest bigmac_run block
capability-auditUser skillcode-backedMedium · ~20-80k

The sprint-audit model generalized to every capability that can be judged from data we already hold. A per-capability spec declares the account data sources (Notion DBs + Fireflies); the model judges each atomic test Pass / Fail / N/A with a 0..1 confidence and verbatim evidence; and the engine emits the standard results file - but only judgments at or above the threshold (default 0.75). Below-bar tests are held and stay In progress for the soft audit, so automation only fills what the data clearly supports.

Read the full guide & setup
Reference - what it brings, its process, and what it needs
BringsMore of the framework audited automatically, honestly - each auto result carries its confidence and an evidence pointer, and capabilities with no central data stay In progress rather than being guessed.
Process
  1. Prepare: resolve the capability's tests + its data sources into a brief.
  2. Gather: pull the declared Notion DBs + Fireflies for the window.
  3. Judge: Pass / Fail / N/A + confidence + evidence per test.
  4. Emit: gate by confidence, write the results file the run merges.
NeedsNotionFirefliesPython bigmac-runmanifest bigmac_run block
sprint-auditUser skillv3.0 · code-backedHeavy · ~80-300k

Evaluates Agile ceremonies against the Big Mac Best Practices Framework - the 47-criterion Scrum Management capability. The model emits a level per criterion (never a number); a deterministic Python orchestrator does all scoring and renders the HTML. Also re-audits single ceremonies and regenerates reports.

Read the full guide & setup
When to use it
run a sprint auditaudit ceremoniesregenerate the sprint reportsre-audit a single ceremonygenerate the client report
How it works
  1. Resolve paths from the manifestReads the renderer, data, and report folders from the account's sprint_audit block.
  2. Prepare transcripts and summariesResolves the engagement, fetches the Fireflies transcripts, and generates the ceremony summaries.
  3. Evaluate each ceremonyJudges each ceremony against the Big Mac criteria and emits a level plus a narrative per ceremony.
  4. Score and validateThe scorer computes the numeric scores deterministically, applies caps, and validates every narrative against the schema.
  5. Render the reportsWrites the synthesis and renders the team, client, internal, and board HTML reports into the report folder.
Example
You say
"run a sprint audit for Acme Robotics / Platform" - a sprint's worth of ceremony transcripts in Fireflies.
You get back
Team, client, internal, and board HTML reports written to the report folder, plus a stdout summary with the overall score and a per-ceremony breakdown (coverage and confidence).

Gotcha: The model emits a level only - every number is computed deterministically by the scorer, never guessed.

Gotcha: It is read-only on the Fireflies transcripts; a missing transcript skips that ceremony and is noted in the scope limitations.

Reference - what it brings, its process, and what it needs
BringsRepeatable, defensible sprint quality scoring with confidence / evidence tracking, cap-if-absent ceilings and repeat-finding escalation - plus ready-to-send client / board / internal reports.
Process
  1. Prepare: orchestrator finds the sprint, loads transcripts, writes per-ceremony summaries.
  2. Evaluate (model): emit a level + narrative per ceremony.
  3. Score: deterministic scorer with coverage + caps; optional second-pass review.
  4. Synthesis & finalize: headline + findings; render all HTML reports.
NeedsFirefliesPython sprint_audit pkgRead/Write/Bashmanifest sprint_audit block