A command that loads likely-needed context into the cache before work begins, based on the user's role, the time of day, and scheduled sprint activities. A cache stores reusable context for later requests.

Use it to prepare role-specific project information for activities such as stand-ups, focused work, backlog refinement, or reporting.

It aims to reduce the wait and repeated loading when common work starts, such as a daily meeting or backlog review.

cache-warm — Command for Claude Code by gonzalezpazmonica

Install

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.

Clone the repo

git clone --depth 1 https://github.com/gonzalezpazmonica/savia

Made for: Claude Code.

Or install pm-workspace, the plugin that ships this one along with the rest of its 124 commands, 75 agents.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/commands/gonzalezpazmonica/savia/cache-warm.svg)](https://agentmods.dev/commands/gonzalezpazmonica/savia/cache-warm)

Your own site

<a href="https://agentmods.dev/commands/gonzalezpazmonica/savia/cache-warm"><img src="https://agentmods.dev/badge/commands/gonzalezpazmonica/savia/cache-warm.svg" alt="Measured on agentmods" height="20"></a>

Per session 17 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 832 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00017	$0.00832
Opus 5	$0.00009	$0.00416
Sonnet 5	$0.00003	$0.00166
Haiku 4.5	$0.00002	$0.00083

Measured yesterday against content hash f15079fe1e3a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

cache-warm 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 yesterday.

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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

cache-warm — 100% identical, 0 lines differ

.claude/commands/cache-warm.md · 126 lines

How it starts

The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/cache-warm

🦉 Savia pre-calienta el caché de forma predictiva — por hora, por rol, por sprint.

Cargar perfil de usuario

Grupo: Context Engineering — cargar:

identity.md — rol del usuario
workflow.md — hora de actividad, eventos del sprint

Ver docs/rules/domain/context-map.md.

Parámetros

--role              Pre-calentar contexto del rol actual
--time auto|HH:MM   Asumir hora (auto = ahora, o especificar hora)
--lang es|en        Idioma del output

Flujo

Paso 1 — Detectar hora y rol

Leer hora actual o parámetro --time
Leer rol de identity.md (PM, Tech Lead, Developer, QA, etc.)
Leer workflow.md para saber si hay eventos hoy (sprint-planning, retro, etc.)

Paso 2 — Mapear contexto probable por hora

Morning (08:00-10:30)

Daily standup context → sprint status, team workload
Load: workflow.md (rutina diaria), team status

Late Morning (10:30-12:00)

Focus time → project context
Load: proyecto activo, specs, decisions

Afternoon (13:00-15:00)

Refinement/grooming
Load: backlog, reglas negocio, estimación

End of Day (15:00-17:00)

Reporting → DORA, KPIs
Load: histórico sprints, métricas

Friday PM (16:00-18:00)

Sprint review + retro
Load: sprint items, decision-log, retrospective patterns

Paso 3 — Mapear contexto por rol

PM/Scrum Master

Always: workflow.md, sprint status, team capacity
If daily: standup context
If friday: review + retro context
If wednesday: backlog grooming context

Tech Lead

Always: tech-radar, team skills matrix
If code-focused day: PR reviews, specs, architecture
If friday: incident postmortems

Developer

Always: my-sprint, my-focus (current task)
If morning: standup, task context
If afternoon: refocus on task, code patterns

QA Engineer

Always: qa-dashboard
If pre-release: test plans, bug triage
If regression day: test cases, known issues

Paso 4 — Ejecutar warm-up

Identificar ficheros a cargar (máx 5 fragmentos)
Cargar fragmentos en caché (sin mostrar contenido)
Registrar warm-up en context-tracking
Mostrar banner:

Read the full file on GitHub · 126 lines

Changes

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.

yesterday First seen · 126 lines · 17 tokens per session scan A f15079fe1e3a

Subscribe to this mod's changes

cache-warm is a command published in the GitHub repository gonzalezpazmonica/savia (50 stars, last pushed yesterday), licensed MIT. It adds 17 tokens to every session and 832 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 2026-09-06.

Memory and context Project management codex harness opencode

implement-plan

Implement the plan as specified, it is attached for your reference. Do NOT edit the plan file itself. Create todos for the plan as you work, starting with the first one. Don't stop until you have completed all the to-dos.

Arenukvern/mcp_flutter · 0 tokens

finish-work

A project wrap-up command for the auto-embedded framework, which supports structured embedded-software development. It runs automated checks, guides a three-part review, records reusable lessons and session notes, and archives the finished task.

DunCanYounG-1/auto-embedded · 0 tokens

handoff

Create a handoff document for seamless session continuity.

travisjneuman/.claude · 0 tokens

talent-motivation

Build a per-person motivation map from observed behavior using Self-Determination Theory + dopamine-anticipation analysis. Output is a motivation hypothesis plus 1-2 leverage interventions, never a personality profile. Refuses engagement-survey-as-data; refuses generic "you should be more motivating." Sub-system 5 of…

frankxai/Starlight-Intelligence-System · 77 tokens

culture-values-ops

Operationalize declared values into systems. For each value, build the matrix — hire criteria, promote criteria, celebration triggers, fire criteria, measurement metric, pay structure — and run the operational test (can an outsider deduce this value from these systems alone?). Refuses values-as-aspiration; insists on…

frankxai/Starlight-Intelligence-System · 71 tokens

chain-load

Loads the most recent chain link to continue work from a previous session.

DeevsDeevs/agent-system · 0 tokens