pi-hypercharm-provider 1.1.5 → 1.2.1

package/.pi/fabric/mesh/state.json

@@ -0,0 +1,20 @@
+{
+  "format": 1,
+  "entries": {},
+  "versions": {
+    "topology/hosts/e9758aa8c95305a8d60a8aa2c20fd53195591d479b2460e1273457793d3f1fec": 18,
+    "sessions/019fe2c8-1dd2-7435-9fa1-fa664831df18": 17,
+    "topology/participants/e9758aa8c95305a8d60a8aa2c20fd53195591d479b2460e1273457793d3f1fec": 18,
+    "topology/hosts/e0efb9d647e046c355b8bff12785e7805d774e7b89892edb38b4b29667ddc15f": 632,
+    "sessions/019fe2d3-1ac4-7f63-a2e7-b396c7bb1738": 631,
+    "topology/participants/e0efb9d647e046c355b8bff12785e7805d774e7b89892edb38b4b29667ddc15f": 632
+  },
+  "tombstoneOrder": [
+    "sessions/019fe2c8-1dd2-7435-9fa1-fa664831df18",
+    "topology/participants/e9758aa8c95305a8d60a8aa2c20fd53195591d479b2460e1273457793d3f1fec",
+    "topology/hosts/e9758aa8c95305a8d60a8aa2c20fd53195591d479b2460e1273457793d3f1fec",
+    "sessions/019fe2d3-1ac4-7f63-a2e7-b396c7bb1738",
+    "topology/participants/e0efb9d647e046c355b8bff12785e7805d774e7b89892edb38b4b29667ddc15f",
+    "topology/hosts/e0efb9d647e046c355b8bff12785e7805d774e7b89892edb38b4b29667ddc15f"
+  ]
+}

package/AGENTS.md

@@ -18,7 +18,8 @@ When a model needs overrides, new properties, or corrections, edit the appropria
 |------|---------|
 | `patch.json` | Per-model overrides keyed by model ID. Add reasoning flags, compat settings, pricing corrections, thinking level maps, etc. Applied on top of `models.json` at runtime and for README generation. |
 | `custom-models.json` | Models that don't exist in the provider API (hidden models, router endpoints, cross-provider aliases). Merged after patch. Format: array of full model objects (same schema as `models.json` entries). |
-| `index.ts` | Provider extension code. |
+| `index.ts` | Provider extension code: model sync, streaming wrapper, footer-status wiring. |
+| `status.ts` | Footer-status presentation: config schema, hypercredit/rate-limit formatters, progressive-disclosure tiers, width-aware widget. Pure module — no pi imports; exercised by `tests/status.smoke.ts`. |
 | `scripts/update-models.js` | The sync script itself (edit only if changing how models are fetched/transformed). |
 
 ## Data Flow

package/README.md

@@ -27,22 +27,21 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
 | Model | Type | Context | Max Tokens | Input Cost | Output Cost |
 |-------|------|---------|------------|------------|-------------|
 | DeepSeek V4 Flash | Text | 1.0M | 384K | $0.20 | $0.40 |
-| DeepSeek V4 Flash 0731 | Text | 1.0M | 384K | $0.15 | $0.30 |
+| DeepSeek V4 Flash 0731 | Text | 1.0M | 384K | $0.20 | $0.40 |
 | DeepSeek V4 Pro | Text | 1.0M | 384K | $2.40 | $4.80 |
-| Gemma 4 26B A4B | Text | 256K | 26K | $0.13 | $0.43 |
-| GLM-5 | Text | 203K | 20K | $0.85 | $2.62 |
-| GLM-5.1 | Text | 203K | 131K | $1.52 | $4.79 |
-| GLM-5.2 | Text | 1.0M | 128K | $1.40 | $4.40 |
+| Gemma 4 26B A4B | Text | 256K | 26K | $0.13 | $0.42 |
+| GLM-5.1 | Text | 203K | 3K | $1.52 | $4.79 |
+| GLM-5.2 | Text | 1.0M | 33K | $1.40 | $4.40 |
 | gpt-oss-120b | Text | 131K | 13K | $0.19 | $0.70 |
-| Kimi K2.5 | Text | 262K | 26K | $0.56 | $2.82 |
-| Kimi K2.6 | Text + Image | 262K | 262K | $0.95 | $4.00 |
+| Kimi K2.5 | Text | 262K | 26K | $0.53 | $2.80 |
+| Kimi K2.6 | Text + Image | 262K | 26K | $0.95 | $4.00 |
 | Kimi K2.7 Code | Text + Image | 256K | 16K | $0.95 | $4.00 |
 | Kimi K3 | Text + Image | 1.0M | 131K | $3.27 | $16.33 |
-| Llama 3.3 70B Instruct | Text | 128K | 13K | $0.51 | $1.04 |
-| Llama 4 Maverick 17B 128E Instruct FP8 | Text | 430K | 43K | $0.28 | $0.93 |
-| MiniMax M2.7 | Text | 205K | 20K | $0.44 | $1.72 |
+| Llama 3.3 70B Instruct | Text | 128K | 13K | $0.60 | $0.74 |
+| Llama 4 Maverick 17B 128E Instruct FP8 | Text | 430K | 43K | $0.27 | $0.90 |
+| MiniMax M2.7 | Text | 262K | 7K | $0.47 | $1.76 |
 | MiniMax M3 | Text + Image | 512K | 512K | $0.33 | $1.31 |
-| Qwen3 Coder 480B A35B Instruct INT4 Mixed AR | Text | 106K | 11K | $0.57 | $2.13 |
+| Qwen3 Coder 480B A35B Instruct INT4 Mixed AR | Text | 106K | 11K | $0.45 | $2.15 |
 | Qwen3 Next 80B A3B Instruct | Text | 262K | 26K | $0.12 | $1.14 |
 | Qwen3.6-Flash | Text + Image | 1.0M | 64K | $1.00 | $4.00 |
 | Qwen3.6-Max | Text | 256K | 64K | $2.00 | $12.00 |
@@ -50,6 +49,7 @@ _Hyperoptimized coding models — DeepSeek, GLM, Kimi, Qwen, MiniMax, Gemma, GPT
 | Qwen3.7-Flash | Text + Image | 1.0M | 64K | $0.20 | $0.80 |
 | Qwen3.7-Max | Text | 1.0M | 64K | $2.50 | $7.50 |
 | Qwen3.7-Plus | Text + Image | 1.0M | 64K | $1.20 | $4.80 |
+| Qwen3.8-Max | Text + Image | 1.0M | 66K | $2.00 | $6.00 |
 *Costs are per million tokens. Prices subject to change — check [hyper.charm.land](https://hyper.charm.land) for current pricing.*
 
 ## Installation
@@ -125,6 +125,51 @@ Supported levels are model-specific and come from Charm's `/v1/provider` catalog
 
 DeepSeek V4 models use the `deepseek` thinking format — the same native format as the [pi-deepseek-provider](https://github.com/monotykamary/pi-deepseek-provider). This sends `thinking: {type: "enabled/disabled"}` plus `reasoning_effort` mapped via `thinkingLevelMap` (`high` → `"high"`, `max` → `"max"`). Replayed assistant messages include empty `reasoning_content` as required by DeepSeek's API.
 
+## Footer Status
+
+A Neuralwatt-style status line sits below the editor. It appears after the
+session's first HyperCharm turn completes (never before — no half-empty line
+on fresh sessions or other providers), refreshes its balance when the agent
+run fully settles, and makes no status-related API calls in sessions that
+never use HyperCharm:
+
+```
+⚡ 1.24 hc · 7 req                       Xu's Team ◆ 249 hc · 996/1k/h · 29d
+└─ session spend+requests ─┘           └─ team · balance · rate limits ──┘
+```
+
+The left side tracks what the current session has sent/spent, read from
+Hyper's `usage.cost.hypercredits` extension on each response (no polling).
+The right side shows the team name (`/v1/teams` — works for API-key auth),
+the Hypercredit balance (`/v1/credits`), the per-hour and per-day request
+rate limits captured from response headers, and days until the OAuth device
+session expires when signed in with OAuth. The right side compresses
+progressively as the terminal narrows, and turns to a warning color at/below
+the `lowBalanceHc` threshold.
+
+### Configuration
+
+Edit `~/.pi/agent/extensions/hypercharm.json` or run `/hypercharm-status`:
+
+| Setting | Values | Default |
+|---------|--------|---------|
+| `session` | `widget` \| `statusbar` \| `off` | `widget` |
+| `account` | `widget` \| `statusbar` \| `off` | `widget` |
+| `hideOnOtherProvider` | `true` \| `false` | `true` |
+| `lowBalanceHc` | number \| `null` | `25` |
+
+Non-interactive toggles:
+
+```
+/hypercharm-status session widget|statusbar|off
+/hypercharm-status account widget|statusbar|off
+/hypercharm-status hide true|false
+/hypercharm-status lowBalance 50|off
+/hypercharm-status refresh
+/hypercharm-status reset
+```
+
+
 ## Authentication
 
 The HyperCharm API key can be configured in multiple ways (resolved in this order):

package/deprecated-models.json

@@ -1 +1,34 @@
-{}
+{
+  "glm-5": {
+    "id": "glm-5",
+    "name": "GLM-5",
+    "reasoning": true,
+    "thinkingLevelMap": {
+      "off": "off",
+      "minimal": null,
+      "low": null,
+      "medium": null,
+      "high": null,
+      "xhigh": null,
+      "max": "max"
+    },
+    "input": [
+      "text"
+    ],
+    "cost": {
+      "input": 0.85,
+      "output": 2.742,
+      "cacheRead": 0.425,
+      "cacheWrite": 0
+    },
+    "contextWindow": 202752,
+    "maxTokens": 20275,
+    "compat": {
+      "supportsStore": false,
+      "supportsReasoningEffort": false,
+      "thinkingFormat": "deepseek",
+      "maxTokensField": "max_tokens"
+    },
+    "deprecatedAt": "2026-08-08T02:00:08.813Z"
+  }
+}