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 agentmods add skills/ashkaan/contextium/projectnpx skills add Ashkaan/contextium --skill projectgit clone --depth 1 https://github.com/Ashkaan/contextiumWhat 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 | $0.00130 | $0.05786 |
| Opus 5 | $0.00065 | $0.02893 |
| Sonnet 5 | $0.00026 | $0.01157 |
| Haiku 4.5 | $0.00013 | $0.00579 |
Grade A, and why
project 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 2d 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 — 282 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/project — the project entry point
/project is the single entry point for project work. It composes the loop's three verbs (think → do → wrap) into one skill that picks the right action based on what you asked and what stage the project is in.
Critical
- Goal-alignment ALWAYS comes first when the think flow runs (per
@rule:depth-policy). The think flow'sthink-step-0-goal-alignmentproduces (goal in plain language + simplest mechanism + explicit ask "does this match?") and MUST wait for explicit user approval before context-load, design, or SPEC writing. Silence is NOT approval. Skipping this is the most expensive mistake the loop can make — a goal/shape mismatch caught up front costs one turn; caught after the work is built it costs the rebuild. - No copy-paste-command outputs as the default. When the router lands on
ready-to-implementin a short session, INVOKE/implement [slug]inline — do not tell the user to paste a command. The only acceptable copy-paste output is the long-session fresh-tab handoff (seestep-2-turn-count-gate), and only because/implement's fresh-context boundary is load-bearing. - Stage detection is a script result, not a judgment. Run
scripts/detect-stage.shto get the stage; don't infer from prose. - Blank-mode: paste the generator output INTO your response. After
step-0.5-render-indexruns Bash, the captured stdout is in your context but NOT yet visible to the user (the Bash tool card is truncated). You MUST copy the entire stdout into your assistant message body. "Ran the generator" is not the deliverable; the user reading the index in your reply IS. - The SPEC stays lean, and
/projectdoesn't write it —/specdoes. The think flow ends by dispatching/spec(think-step-4-dispatch-spec), which writes the four-section SPEC (Ask / Behavior / Files / Done per.claude/templates/spec-lean.md), spirit-checks it, and auto-closes (commits it) — the user reviews the committed SPEC in the fresh/implementtab, not at a sign-off gate. Don't pad it with boilerplate that degrades to "N/A"; a heavy project grows its own sections when a real gap bites, per@rule:simplest-solution-default.
What ships with it
8 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.
- references/readme-template.md 2.0 KB
- references/spec-schema.md 2.7 KB
- references/think-flow-deep.md 5.2 KB
- scripts/check-staleness.sh 6.2 KB runs code
- scripts/check-staleness.test.sh 7.1 KB runs code
- scripts/detect-stage.sh 3.5 KB runs code
- scripts/find-project.sh 1.5 KB runs code
- scripts/parse-arg-mode.sh 1.8 KB runs code
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.
- 2d ago First seen · 282 lines · 130 tokens per session scan A ee7791383dc9
project is a skill published in the GitHub repository Ashkaan/contextium (32 stars, last pushed 5d ago), licensed MIT. It adds 130 tokens to every session and 5,786 once invoked, about $0.0006 per session on Opus 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 2026-08-30.
Other skills, from other repositories
sovereign-economics-engine
Unified business model evaluator, client filter, pricing engine, and distribution architect. Absorbs 31 business + 8 marketing + 13 content + 2 acquisition protocols.
therapeutic-ifs
Unified inner work engine: Schema deconstruction (diagnosis) + IFS therapy (treatment). Absorbs: schema-deconstruction.
dashboard-builder
Build self-contained interactive HTML dashboards with charts, filters, and tables. Generates a single browser-openable file — no server or dependencies required.
bionic-decision-engine
Unified mathematical arbitrator for all resource allocation decisions — money, time, energy, relationships. Absorbs 46 decision protocols + 24 strategy protocols into one dense engine.
statistical-analysis
Structured pipeline for statistical analysis deliverables — SPSS, R, Python. Covers reliability, chi-square, correlation, regression, assumption checking, and client-ready reporting.
decision-journal
Unified decision lifecycle: Pre-decision logging, post-decision review, failure classification, and calibration tracking. Absorbs: post-mortem-engine.