Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add jeffreytse/grimoire-core --skill apply-pdcagit clone --depth 1 https://github.com/jeffreytse/grimoire-coreWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jeffreytse/grimoire-core/apply-pdca)<a href="https://agentmods.dev/skills/jeffreytse/grimoire-core/apply-pdca"><img src="https://agentmods.dev/badge/skills/jeffreytse/grimoire-core/apply-pdca/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/jeffreytse/grimoire-core/apply-pdca"><img src="https://agentmods.dev/badge/skills/jeffreytse/grimoire-core/apply-pdca.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00034 | $0.01444 |
| Opus 5.5 | $0.00014 | $0.00578 |
| Sonnet 5.5 | $0.00007 | $0.00289 |
| Haiku 4.5 | $0.00003 | $0.00144 |
Grade A, and why
apply-pdca scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Apply PDCA
Run a disciplined four-phase cycle — Plan, Do, Check, Act — to make improvements systematically, learn from each iteration, and lock in gains before scaling.
Why This Is Best Practice
Adopted by: ISO 9001 (adopted by 1.1 million organizations in 170 countries as of 2022), Toyota Production System (Kaizen), Six Sigma (DMAIC is a direct extension), NHS (UK National Health Service quality improvement), U.S. Department of Defense, and standard curriculum in every accredited operations management program globally. Impact: Toyota's application of PDCA-based Kaizen is directly credited with its position as the world's most efficient major automaker — Toyota's defect rate is consistently 50–70% lower than industry average (J.D. Power IQS longitudinal data). ISO 9001 meta-analyses show certified organizations achieve 15–30% reduction in defect rates and measurable customer satisfaction improvement within two years of adoption (Psomas & Antony, 2015, International Journal of Quality & Reliability Management). Why best: Most improvement attempts skip the Check phase — they implement a change and assume it worked. PDCA enforces measurement of actual results against predicted results before locking in or scaling the change. This prevents three failure modes: improvements that don't actually work, improvements that work but create side-effects, and improvements that are scaled before they're understood.
Sources: W. Edwards Deming "Out of the Crisis" (1986); Shewhart "Statistical Method from the Viewpoint of Quality Control" (1939); ISO 9001:2015 §10.3; Psomas & Antony (2015) IJQRM
Steps
- Plan — Define the problem — Write the current state in measurable terms. "Customer support tickets take an average of 48 hours to resolve; target is 24 hours." Vague problem statements produce vague plans.
- Plan — Identify root cause — Use Five Whys or fishbone analysis to find the systemic cause before designing the fix. Fixing symptoms produces the symptom again within weeks.
- Plan — Design the change — Write the specific intervention: what will change, who owns it, what resources are needed, and what the predicted outcome is in measurable terms. The prediction is what the Check phase validates.
- Plan — Define success metrics — Choose 1–3 measurable indicators that will tell you whether the change worked. Set a target and a timeline for measurement.
- Do — Run a small-scale pilot — Implement the change in a limited scope: one team, one shift, one product line, or one week. Never scale an untested change organization-wide. Document everything that happens, including deviations from the plan.
- Check — Measure actual vs. predicted — At the end of the pilot, compare actual results to the predicted outcome from step 3. Did the metric move? By how much? Were there unexpected side effects?
- Check — Identify what you learned — If results match prediction: the change works. If not: determine whether the root cause was wrong, the intervention was wrong, or the measurement was wrong. Each answer leads to a different next cycle.
- Act — Standardize or abandon — If the change worked: update the standard operating procedure, train all affected people, and remove the old way. If it didn't: return to Plan with updated understanding. Never leave a "pilot" running indefinitely without deciding to standardize or discontinue.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 58 lines · 34 tokens per session scan A 75621fe1f0d3
apply-pdca is a skill published in the GitHub repository jeffreytse/grimoire-core (5 stars, last pushed 5d ago), licensed MIT. It adds 34 tokens to every session and 1,444 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
semantic-model-disambiguation
Analyze Power BI semantic models for column-level overlaps that confuse Copilot and Fabric data agents. Detect ambiguity, review with domain expert, apply fixes via MCP.
prompt-2-data
Generate comprehensive synthetic relational data for any specified subject with multiple normalized CSV files maintaining referential integrity.
academic-paper-drafting
End-to-end academic paper drafting for CHI, HBR, journals, and conferences with venue-specific templates, drafting workflows, and revision strategies.
research-first-development
Build knowledge bases that build software — research before code, teach before execute.
infrastructure-as-code
Domain: DevOps & Cloud Engineering.
agent-governance
Patterns for adding safety, trust, and policy enforcement to AI agent systems -- control which tools agents can call, what content they process, and maintain accountability through audit trails.