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/decrec-io/decrec-plugin/decrecnpx skills add decrec-io/decrec-plugin --skill decrecgit clone --depth 1 https://github.com/decrec-io/decrec-pluginWhat 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 | $0.00143 | $0.06810 |
| Opus 5 | $0.00072 | $0.03405 |
| Sonnet 5 | $0.00029 | $0.01362 |
| Haiku 4.5 | $0.00014 | $0.00681 |
Grade A, and why
decrec 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 — 411 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pick a Route
Match the most specific row. Don’t advance to a later stage (shape/update/explore/admit/commit/decline/archive) unless they asked for it. Clarifying a thin brief during Shape / Explore is part of that stage — not a one-shot requirement.
If this chat is Decision-worthy and they didn’t name a DEC-n, run Recall first, then the matched row. Ordinary coding, bugfixes, and implementing an already-chosen approach do not qualify.
Tip = the current brief version (briefVersion + brief from response). Update, Explore-existing, Follow-up, Commit, and Decline target that tip.
Pins = briefVersion + expectedDecisionVersion from response. Commit and Decline also need explorationRunId from response recommendation. Archive needs expectedDecisionVersion only. Always load decision for pins. If this chat already has a briefVersion for that DEC (create, update, or a prior load), a later load is a reload: compare only that number. Unchanged → do not inspect brief text. Changed → stop, show the six brief-field changes, ask. Do not admit, commit, decline, or update-from-stale onto the new tip.
Find without DEC-n: Topic in the ask (“why / what decisions about…”) → search_decisions (not list). Recency / what’s open / pick among recent with no topic → list_decisions. One clear hit → use it. Several → show displayId + title (+ stage/snippet for search) and ask. Don’t invent. Search hits are not pins — still get_decision before update/admit/commit/decline/archive. matchQuality: "weak" → say nothing ranked well; still show hits; do not claim the org has no decisions. Do not page list_decisions hoping titles match a theme. If nextCursor is set on list and they need more, pass it as cursor.
| Intent | Path |
|---|---|
| Pasted thread / transcript / meeting notes (extract / what to record) | Catalog → stop. Shape only picked rows. Admit / Explore only if that was in the ask |
| Explore from scratch | Shape brief → Explore (Densify, then confirm brief) → stop |
Explore existing DEC-n |
Load decision → stop if locked / not approved / in-flight → Explore on tip (as-is), even if recommendation is set → stop |
| Admit framing only | Shape brief → Create brief-only → stop |
| Admit new + pick | Shape brief + exploration → Create with exploration → stop |
Admit onto existing DEC-n (same brief) |
Load decision (Pins) → stop if locked / not approved / in-flight → Follow-up (Shape exploration if not in hand) → stop |
Update existing DEC-n (brief ± pick) |
Load decision (Pins) → stop if locked / in-flight → Shape from tip + requested changes → Update (include exploration if admitting a pick) → stop |
| Already explored in chat | Shape brief + exploration → Create / Update / Follow-up (skip Explore) → stop |
Load / discuss DEC-n |
Load decision only |
| Recency / pick among recent (no topic) | list_decisions → ask unless one clear hit → Load if they need the brief |
| Why / what decisions about a topic | search_decisions → show ranked hits → ask unless one clear hit → get_decision before update/admit/commit/decline/archive |
Commit existing DEC-n |
Load decision (Pins) → stop if locked / not approved / in-flight / no recommendation → Commit → stop |
Decline existing DEC-n |
Load decision (Pins) → stop if locked / not approved / in-flight / no recommendation → Decline → stop |
Archive existing DEC-n |
Load decision (Pins) → stop if status=archived → Archive → stop |
About to reopen a Decision-worthy direction (no DEC-n) |
Recall → search_decisions → then reason |
| Which account / org this MCP session is | whoami → report email + org → stop. Skip if initialize instructions already named it and they did not ask |
| No DecRec ask; this chat just settled a Decision-worthy fork | Capture → offer once → stop unless they accept |
What ships with it
1 file 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 · 411 lines · 143 tokens per session scan A 626dd295d2a2
decrec is a skill published in the GitHub repository decrec-io/decrec-plugin (1 stars, last pushed 3d ago), licensed MIT. It adds 143 tokens to every session and 6,810 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
aws-lambda-microvms
Build, run, debug, and operate applications on AWS Lambda MicroVMs — Firecracker-isolated, snapshot-resumable serverless compute environments that run inside a container with up to 8-hour lifetimes. Triggers on: Lambda MicroVMs, Firecracker isolation, snapshot-resumable compute, suspend/resume, sandboxed or untrusted…
document-service
This skill should be used when the user asks to "analyze this codebase", "document this service", "generate technical docs", "I inherited this code", "help me understand this system", "create docs for this project", "what does this system look like", "onboard me to this codebase", "this codebase has no docs"…
hyperpod-cluster-debugger
Diagnose and remediate cluster-wide HyperPod (EKS or Slurm) problems — creation / deployment failures (CloudFormation, EFA health check, lifecycle scripts, capacity), EKS access, node replacement, CloudFormation nested-stack errors, post-maintenance rollback state, dangling nodes, autoscaler conflicts. Includes…
hyperpod-nccl
Diagnose NCCL failures and adjacent training-pod failures on HyperPod GPU clusters (EKS or Slurm) — training hangs, AllReduce / collective-op timeouts, EFA or libfabric errors, rendezvous failures, EFA TCP fallback, /dev/shm or memlock issues, NCCL version mismatch across pods, container OOM / exit-137 / OOMKilled…
planning
Discovers user intent and generates a structured, step-by-step plan for model customization workflows. This skill must always be activated alongside any other skill when the user's request relates to model customization — including fine-tuning, training, building, customizing, reviewing data, or getting advice on…
hyperpod-issue-report
Generate comprehensive issue reports from HyperPod clusters (EKS and Slurm) by collecting diagnostic logs and configurations for troubleshooting and AWS Support cases. Use when users need to collect diagnostics from HyperPod cluster nodes, generate issue reports for AWS Support, investigate node failures or…