A command that groups Claude usage into five-hour session windows. It shows how usage falls within each window, using recorded reset times when available and estimates when they are not.

It helps inspect five-hour usage blocks, sort their cost and cache details, and compare the results with compatible usage reports.

It makes short-term usage and rate-limit windows easier to understand while showing when a boundary is only an estimate.

blocks — Command by omrikais

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.

agentmods

npx agentmods add commands/omrikais/cctally/blocks

Clone the repo

git clone --depth 1 https://github.com/omrikais/cctally

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 2,468 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

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	$0.00000	$0.02468
Opus 5	$0.00000	$0.01234
Sonnet 5	$0.00000	$0.00494
Haiku 4.5	$0.00000	$0.00247

Measured 2d ago against content hash a57a1e62bbb5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

blocks 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.

docs/commands/blocks.md · 163 lines

How it starts

The opening of the file, as written. The whole thing — 163 lines — stays where its author put it; the contents beside it link to each section on GitHub.

`blocks`

Claude usage grouped by 5-hour session block.

Cost coverage: Claude dollar and token totals are transcript-derived lower bounds, not exact /usage billing totals.

Canonical form: cctally claude blocks (this flat form remains as an alias).

Which 5h command? blocks (this page) is the ccusage-compatible per-window view (includes ~-prefixed heuristic rows when no recorded reset covers a window); five-hour-blocks is the API-anchored analytics view with model/project rollups and 7d drift; five-hour-breakdown drills into per-percent milestones inside one block.

cctally blocks aims for parity with ccusage blocks's output shape but implements two intentional improvements over upstream:

5h block boundaries follow Anthropic's rate_limits.5h.resets_at (10-minute granularity) wherever a recorded snapshot covers the window, not the legacy hour-floor algorithm. A first message at 7:10 PM anchors the block at 7:10, not 7:00. ccusage still uses floor_to_hour (rust/crates/ccusage/src/blocks.rs:51-54). Windows with no recorded reset fall back to a heuristic start and render as ~-prefixed rows that can drift from the real boundary (see Notes).
Duplicate-pair dedup matches ccusage's should_replace_deduped_entry (rust/crates/ccusage/src/claude_loader.rs:531) — higher-token-total wins, speed-set breaks ties. Fixed in v1.12.0; pre-fix versions kept the streaming-intermediate row and systematically under-counted output tokens on tool-using turns.
Block projection uses the real reset start, not entry-span. cctally projects tokens ÷ (now − real reset start) × full 5h window; ccusage uses current + (tokens ÷ (last entry − first entry)) × time remaining (projectBlockUsage). cctally's model leverages the API-anchored window start that ccusage only guesses, so the Projected Usage numbers differ by design. The -a box and the table's PROJECTED footer both reflect this formula.

Read the full file on GitHub · 163 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.

2d ago First seen · 163 lines · 0 tokens per session scan A a57a1e62bbb5

Subscribe to this mod's changes

blocks is a command published in the GitHub repository omrikais/cctally (5 stars, last pushed 3d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 2,468 tokens. 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.

Analytics and metrics Writing and docs anthropic ccusage claude claude-code cli codex

blocks

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

`blocks`

What this file has done since we first saw it

Other commands, from other repositories

checklist

clarify

specify

analyze

converge

implement