@clemrig/flow-ui 0.8.0 → 0.9.0

package/README.md

@@ -11,7 +11,7 @@ npm install -g @clemrig/flow-ui
 flow-ui [RUNS_DIR] [--port PORT]
 ```
 
-`flow-ui` serves the runs kept in `RUNS_DIR` (default: `.flow/runs`, where `@clemrig/flow` keeps them; a `RUNS_DIR` given must be an existing directory) on `127.0.0.1`, on `PORT` or a port the system chooses, prints its URL and stops on Ctrl-C. The page shows the runs in a rail you can fold, grouped by day, named by their input, filtered by status or workflow and searched (`/`). A run shows its name, status, links its result gives, input and figures (start, end, active time, cost against the initial budget with each budget allotted and where it comes from on hover, then launches, rounds and stage executions); its launches (the first one, then each resumption) on one bar, each sized by its active time with its day and start time, the pause between two written out, and why each failed one stopped, a launch opening what it ran with (versions, prompts, runtimes, budget and its origin, machine, directory, the runtimes' variables set, values given), what changed since the launch before marked; its stage executions over its active time, a lane per agent stage, with the resumptions, the rounds its corrections started and each correction taken, zoomed on demand (Fit, 2×, 4×, or a pinch or Ctrl/⌘ with the wheel on the lanes); its correction counters. Below, the selected execution's panel with four tabs: Overview (outcome, the agent call's model, session and settings, saying which ones the profile leaves to the machine or the runtime, tools, whether the call opens its conversation or continues it, tokens and cost), Input (the correction it received, then what flow sent each call, loaded on demand: the workflow's prompt, the instructions, the context and the output schema as trees, the whole request and the command), Output (its agent's answer, the correction it sent back or could not send, and each call's answer when it made several) and Activity (its calls, tool calls, journal and warnings in one timeline you can filter). Beside it, its cost by session and over its active time with each launch's budget when known, then its result, its data as a tree you unfold, and its runtimes. The address keeps the run, the execution chosen and the tab (`#/<id>/stages/<n>/<tab>`); `j`, `k`, `f` and `1` to `4` move between executions and tabs. It follows a running run as it goes, and follows the selected run again when it is resumed. A result shows the run's validated data as JSON.
+`flow-ui` serves the runs kept in `RUNS_DIR` (default: `.flow/runs`, where `@clemrig/flow` keeps them; a `RUNS_DIR` given must be an existing directory) on `127.0.0.1`, on `PORT` or a port the system chooses, prints its URL and stops on Ctrl-C. The page shows the runs in a rail you can fold, grouped by day, named by their input, filtered by status or workflow and searched (`/`). A run shows its name, status, links its result gives, input and figures (start, end, active time, cost against the initial budget with each budget allotted and where it comes from on hover, then launches, rounds and stage executions); its launches (the first one, then each resumption) on one bar, each sized by its active time with its day and start time, the pause between two written out, and why each failed one stopped, a launch opening what it ran with (versions, prompts, runtimes, budget and its origin, machine, directory, the runtimes' variables set, values given), what changed since the launch before marked; its stage executions over its active time, a lane per agent stage, with the resumptions, the rounds its corrections started and each correction taken, zoomed on demand (Fit, 2×, 4×, or a pinch or Ctrl/⌘ with the wheel on the lanes); its correction counters. Below, the selected execution's panel with four tabs: Overview (outcome, the agent call's model, session and settings, saying which ones the profile leaves to the machine or the runtime, tools, whether the call opens its conversation or continues it, tokens and cost, then the session each call ran in as its runtime confirmed it, where it differs from the profile marked), Input (the correction it received, then what flow sent each call, loaded on demand: the workflow's prompt, the instructions, the context and the output schema as trees, the whole request and the command), Output (its agent's answer, the correction it sent back or could not send, and each call's answer when it made several) and Activity (its calls, tool calls, journal, the hooks that ran, a failed one marked, what the runtime reported, such as refused tools, limits, retries and compactions, and warnings, in one timeline you can filter). Beside it, its cost by session and over its active time with each launch's budget when known, then its result, its data as a tree you unfold, and its runtimes. The address keeps the run, the execution chosen and the tab (`#/<id>/stages/<n>/<tab>`); `j`, `k`, `f` and `1` to `4` move between executions and tabs. It follows a running run as it goes, and follows the selected run again when it is resumed. A result shows the run's validated data as JSON.
 
 To watch a run on another machine, forward the same port through SSH: `ssh -L 4321:127.0.0.1:4321 machine`, run `flow-ui --port 4321` there, and open `http://localhost:4321/`.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clemrig/flow-ui",
-  "version": "0.8.0",
+  "version": "0.9.0",
   "description": "A local read-only dashboard of the runs of @clemrig/flow workflows",
   "license": "MIT",
   "repository": {
@@ -35,7 +35,7 @@
     "prepack": "npm run build"
   },
   "dependencies": {
-    "@clemrig/flow": "0.9.0"
+    "@clemrig/flow": "0.10.0"
   },
   "devDependencies": {
     "@testing-library/dom": "10.4.2",