token-weather-usage
Reviewed
at commit 3c7e878
New version, not yet reviewed
One line above the prompt with context weather, a bar per prompt sized by the tokens it added, and 5-hour and 7-day limit gauges that hatch the gap with…
0 upvotes
Log in to upvote or save- Made by
- Eric Cologni
- Pinned commit
-
3c7e87817c60d4a9f102287cb81d309569e1250c - What we checked
-
Draws one line above the prompt with context size and weather, bars per recent prompt, 5-hour and 7-day limit usage, prompt cache time left, session cost and running subagents. A pill appears when Claude's answer proposes quoted GO approvals; clicking its OK sends them as your message (can be switched off). It makes no network calls and keeps only usage figures locally.
Review details
Static review of the source at the pinned commit; not run. Delta review of the changes since 0fb3819; the rest carries over from the review at that commit. Registers session.start, session.end, turn.step, turn.complete, session.compact, session.measure and ui.render (AbovePrompt). Reads $.session.usage and $.session.id, $.agent.list, token usage and tool names from turn.step results (names only, never arguments), and the environment variables LC_ALL, LC_MESSAGES, LANG, DISABLE_PROMPT_CACHING, FORCE_PROMPT_CACHING_5M, CLAUDE_CODE_PROMPT_CACHE_TTL and ENABLE_PROMPT_CACHING_1H. Stores per-session readings, cache state, cost and limit history in $.store (sessions idle for more than 8 days are deleted). Tool calls and requests are only observed. The GO pill (option goPill, on by default) scans the main agent's last answer for quoted phrases starting with GO and, only when the user presses its OK button, calls $.prompt.submit with 'You have my approval for all the GOs: ...' as the user's own message. Changes in this version: it now also reads the session's model with $.session.model() to warn that the prompt cache does not follow a model switch; a daily-budget pill and an end-of-week estimate for plans that show only the 7-day limit, computed from already stored readings; a longer bar in that case; new labels in English, French and Russian; and price table entries for claude-fable-5-2 and claude-mythos-5-2. No fetch, sockets, processes, file access, eval or dynamic code; no external hosts. README, tests and icon not reviewed beyond file listing.
- What it uses
-
- Reads your prompts: no
- Sees tool calls: yes
- Blocks or changes tool calls: no
- Reads or writes files: no
- Runs programs on your machine: no
- Calls the model (uses tokens): no
- Sends data over the network: no
- Stores data: yes
- Shows something on screen: yes
One line above the prompt: context tokens with their weather, a bar per recent prompt, your 5-hour and 7-day limits against the time elapsed, the prompt cache with its time left, the session cost with the last prompt's share in dollars and in points of the 5-hour limit, and the subagents running. Labels in English or French. After Token Weather (Anthropic, Apache-2.0) and the ideas of usage-meter (HolyGrail) and prompt-cache-control (Daniel San).
Made this mod? Add a badge to your README
Paste this into your README. The badge says reviewed once the current version has been reviewed, and links readers to this page.
[](https://modsforclaude.com/mods/token-weather-usage/)