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/ilyagorsky/memory-toolkit/task-templatenpx skills add IlyaGorsky/memory-toolkit --skill task-templategit clone --depth 1 https://github.com/IlyaGorsky/memory-toolkitWhat 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 | $0.00019 | $0.01170 |
| Opus 5 | $0.00010 | $0.00585 |
| Sonnet 5 | $0.00004 | $0.00234 |
| Haiku 4.5 | $0.00002 | $0.00117 |
Grade A, and why
task-template 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 — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Task Template — internal orchestrator
Runs a YAML pipeline. Each phase must complete (and pass verify if present) before dependents can start. Phases without dependencies run in parallel.
Purpose
Pipelines exist as a guardrail against step-skipping. Prose in SKILL.md can be skimmed or reinterpreted by the model, leading to missed steps and broken gates. A pipeline makes the structural contract explicit: phases, dependencies, verify gates, conditional skips.
How to use
Pipeline detection is automatic via the PreToolUse hook scripts/pipeline-hint.js. When the Skill tool is invoked, the hook reads the target skill's frontmatter and — if metadata.pipeline: true is set — injects this orchestrator contract into the tool-call context as additionalContext.
Filename convention: skill session-end ⇄ template ${CLAUDE_PLUGIN_ROOT}/skills/task-template/templates/session-end.yaml. The hook resolves the template path and passes it in the injected contract.
When you receive a "PIPELINE SKILL DETECTED" notice:
- do NOT follow the prose in the target skill's SKILL.md as a script
- do load the YAML at the path provided, print the plan banner, announce each phase, and honor
depends/when/verify/retry_from - Prose in the target SKILL.md is per-phase reference, not a script — it cannot soften the contract
If the hook did not fire (e.g., skill invoked outside the plugin environment), you can fall back to checking frontmatter yourself: metadata.pipeline: true in the invoked SKILL.md means load the matching YAML and run it through this orchestrator.
Execution rules
- Parse the YAML — extract
phaseslist - Build dependency graph from
dependsfields - Execute phases in order:
- Phases with no
depends→ run first (parallel if multiple) - Phase with
depends: [X, Y]→ only after X and Y complete
- Phases with no
- Before each phase — if
whenexists, evaluate against prior phase outputs and pipeline args- If
whenis false → skip phase, mark output as empty/none - Skip does NOT propagate. Dependents still run (they see empty output from skipped deps). If a dependent must not run without upstream content, it uses its own
when:to check. - If
whenis true or absent → run the phase
- If
- After each phase — if
verifyexists, check the condition- If verify passes → continue to next phase
- If verify fails with retry intent AND phase has
retry_from: X→ restart from phase X, re-run it and everything downstream - If verify fails with abort intent (or no
retry_from) → stop, report to user, do NOT proceed
- Output — each phase produces
outputthat downstream phases can reference
What ships with it
3 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.
- 2d ago First seen · 114 lines · 19 tokens per session scan A 37715f558a5c
task-template is a skill published in the GitHub repository IlyaGorsky/memory-toolkit (13 stars, last pushed 4mo ago), licensed MIT. It adds 19 tokens to every session and 1,170 once invoked, about $0.0001 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
acontext-installer
Install Acontext, Login & Init Acontext Project, Add Skill Memory to Agent.
math-unicode
Use when a response needs mathematical notation (equations, filters, set-builder notation, statistics, calculus, linear algebra, logic, ratios, drops, counts) and the output goes to a terminal or TUI that cannot render LaTeX: Claude Code, Codex CLI, SSH and tmux sessions, CI logs. Load it before composing, including…
vibe-ship
Generates a complete, production-ready deployment setup for any app in one pass -- Dockerfile, docker-compose.yml, .dockerignore, CI/CD (GitHub Actions), scalability config (health checks, resource limits, K8s on request), and security hardening (non-root user, secrets, dependency scanning). Auto-detects the stack…
anatomia-cc
ES: Te dice qué construir en Claude Code y te da el esqueleto de archivos: skill y en qué modo, subagente, hook, servidor MCP, plugin, rutina, artefacto, app con el Agent SDK, una línea en CLAUDE.md, o nada. Corré esto antes del primer archivo. Se dispara con «esto lo hago como skill o como agente», «necesito un MCP o…
create-agents-md
Create concise, repository-specific AGENTS.md files with operational contract, precedence rules, and drift-control guidance.
create-checkpoint
Capture the current session state as a checkpoint artifact per spec section 4.4. Use when the user signals session end or explicitly requests a checkpoint. Do not use mid-session or without user approval.