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 sungurerdim/dev-skills --skill ds-blueprintgit clone --depth 1 https://github.com/sungurerdim/dev-skillsWrote 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/sungurerdim/dev-skills/ds-blueprint)<a href="https://agentmods.dev/skills/sungurerdim/dev-skills/ds-blueprint"><img src="https://agentmods.dev/badge/skills/sungurerdim/dev-skills/ds-blueprint/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/sungurerdim/dev-skills/ds-blueprint"><img src="https://agentmods.dev/badge/skills/sungurerdim/dev-skills/ds-blueprint.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 | $0.06267 |
| Opus 5 | $0.00017 | $0.03134 |
| Sonnet 5 | $0.00007 | $0.01253 |
| Haiku 4.5 | $0.00003 | $0.00627 |
Grade B, and why
ds-blueprint scanned grade B with 1 finding 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 5d 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.
Enumerates other installed skillsmediumAgent snooping
Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.
5. **Skillset stamp.** Read `.dev-skills-version` beside the installed skills dir (e.g. `~/.claude/skills/.dev-skills-version`); absent → `unknown`. How it starts
The opening of the file, as written. The whole thing — 267 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/ds-blueprint
Can't improve what you don't measure. Skill scores project across 9 dimensions, writes the signal inventory every other skill scopes itself by, and tells you exactly where to focus next.
Project Health System — Profile-based assessment, transformation, and progress tracking.
Completion Evidence — applies to every phase: Report
done/OKonly with the machine-checkable evidence the gates name — the exact command run and its observed output (orfile:linediff). Missing evidence → reportINCOMPLETEplus what is missing. Self-assessment is never evidence. (This band repeats at file end by design — both copies are normative.)
Triggers
- User runs
/ds-blueprint - User asks to assess project health/quality/status, or for a profile, health score, or dashboard
- User asks "how healthy is this project" or "what should I improve"
- First time working on new project (suggest profile creation)
- Verification-infrastructure gap (no CI/tests/linter) surfaces during any task (suggest audit, don't auto-invoke)
Triggers — INVOKE / DON'T INVOKE
| INVOKE | DON'T INVOKE |
|---|---|
| "score project health", "what should I improve first" | "fix the issues" (→ ds-review or ds-fix) |
| "create project profile", "refresh blueprint" | "research industry best practices" (→ ds-research) |
| "blueprint dashboard, 9 dimensions" | "release-candidate report" (→ ds-ship) |
| "where is this codebase weakest" | "competitor comparison" (→ ds-benchmark) |
| "project has no CI/tests — what else is missing?" | "scaffold CI from zero on a greenfield project" (→ ds-init) |
Contract
Dimensions: B2, B4 (contributor), A9 (signal)
- Signal counting, not file:line finding lists. Only modifies the profile section of the instruction file; suggests next steps but never invokes other skills or fixes code.
- Standalone. Uses blueprint profile or
ds/audit/findings.mdwhen available; own analysis when absent. Shared formats: ../core/findings-and-profile-format.md; signal keys and detection: ../core/signal-inventory.md; severity and score: ../core/severity-score-categories.md. - Full accounting enforced: every finding and planned check ends in an explicit disposition (fixed / skipped + reason / only you can do); summary totals balance.
- Pre-existing / out-of-scope errors detected during work are NOT skipped — fixed inline or escalated with concrete blocker.
- SSOT (W10):
ds/audit/findings.mdis every fix skill's single source of truth — blueprint MUST detect every issue in-scope (a miss stays unfixed downstream). Consumers (ds-review, ds-fix, …) defer to a fresh copy — fresh =git_hash == HEADAND from this run-cycle (this invocation, or its orchestration run) — verify + apply only, never re-detect blueprint's owned scopes. Stale/missing → orchestrated: invoke/ds-blueprint --refreshand wait; standalone: announcefindings stale — running own {scopes} analysis, scan, append with ownsource+ currentgit_hash(prior-cycle rows = diff baseline only). Re-detecting a covered scope this cycle violates W10; skipping a re-scan for a recent prior cycle violates W11. - Privacy has one owner.
privacyfindings come from ds-compliance (canonical); blueprint consumes them for Security & Privacy when present, else scores fromsecurityalone (dashboard:privacy: not scanned (ds-compliance)). Blueprint never writesprivacyrows. - Overwrite-only persistence: state (
ds/audit/blueprint.json— a full 9-dimension scan holds its progress nowhere else, so an interruption restarts it from zero), findings, and profile rewritten every run — never appended. Run history lives ingit log -- <instruction-file>, not in profile ords/audit/. - Human-action items: findings needing human-only access (branch protection, CI/repo secrets, store/account setup, key rotation, purchases) get a distinct
Human actionsblock in Dashboard + Summary — never dropped, never marked fixed by the AI. - Dev-Value Gate on every profile line (../core/findings-and-profile-format.md § 2, forbidden-line list there): a line earns its place only if it makes AI engineering measurably better for the next 6 months; forbidden classes go to README/CHANGELOG/git log/terminal summary.
What ships with it
9 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 5d ago Changed · -73 lines · -1 tokens per session scan A → B b66c3bd0137c
- 8d ago First seen · 340 lines · 35 tokens per session scan A a505aa02a396
ds-blueprint is a skill published in the GitHub repository sungurerdim/dev-skills (1 stars, last pushed 3d ago), licensed MIT. It adds 34 tokens to every session and 6,267 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (enumerates other installed skills). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
ring:dispatching-workflows
Executing a phased plan in rolling waves where each phase runs as one multi-agent workflow harness: the supervisor elaborates the phase into tasks against the real landed code, launches a workflow that implements with TDD and runs mandatory in-harness review plus an adversarial contrarian pass (and researchers when…
ring:implementing-tasks
Implementing a single planned task (Task N.M.T) end-to-end: selects the right backend agent by language and service type, drives one TDD RED->GREEN turn, then verifies coverage, lint, license headers, runtime, and delivery before handoff. Runs as Gate 0 before ring:reviewing-code. Use to drive ONE task inside an…
ring:managing-dev-cycle
Managing an in-progress development cycle without driving it: status reports phase, epic/gate progress, assertiveness, and elapsed time from current-cycle.json; cancel confirms, marks the cycle cancelled, and writes a partial feedback report. Use when checking the status of, or cancelling, a running dev cycle. Skip…
ring:mapping-feature-relationships
Mapping how features relate and phasing the work: categorizing PRD features, grouping them into domains, charting cross-feature journeys, dependencies, and integration points, and defining the binding Phases that plan.md mirrors one-to-one at Gate 7. Gate 2 of ring:planning-large-features; runs after…
ring:planning-large-features
Planning the 8-gate Large Track pre-dev workflow (research, PRD, feature map, TRD, API contract, data model, dependency map, plan) with per-gate human approval. Use for features 2+ days that add dependencies, data models, multi-service integration, or new architecture. Skip for small features (use…
ring:planning-small-features
Planning the lightweight 4-gate Small Track pre-dev workflow (research, PRD, TRD, plan) with human approval and state tracking at each gate. Use for features under 2 days that reuse existing patterns and add no new dependencies, data models, or services. Skip for larger or complex features — use…