roadmodel 0.2.94__tar.gz → 0.2.96__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {roadmodel-0.2.94 → roadmodel-0.2.96}/PKG-INFO +1 -1
- {roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs}/catalog.json +1 -1
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/claude-commands/roadmap-refresh.md +33 -31
- {roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/claude-commands}/roadmap-step.md +17 -8
- {roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/claude-commands}/roadmodel-upgrade.md +19 -1
- {roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/templates}/phase-roadmap-template.md +91 -70
- {roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/templates}/planning-kit-how-to-use.md +2 -2
- {roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/templates}/project-roadmap-template.md +72 -53
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/templates/prompt-phase-roadmap.md +8 -8
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/templates/prompt-project-roadmap.md +1 -1
- {roadmodel-0.2.94 → roadmodel-0.2.96}/pyproject.toml +1 -1
- {roadmodel-0.2.94/docs → roadmodel-0.2.96/roadmodel/data}/catalog.json +1 -1
- {roadmodel-0.2.94/docs/templates → roadmodel-0.2.96/roadmodel/data}/phase-roadmap-template.md +91 -70
- {roadmodel-0.2.94/docs/templates → roadmodel-0.2.96/roadmodel/data}/planning-kit-how-to-use.md +2 -2
- {roadmodel-0.2.94/docs/templates → roadmodel-0.2.96/roadmodel/data}/project-roadmap-template.md +72 -53
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/prompt-phase-roadmap.md +8 -8
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/prompt-project-roadmap.md +1 -1
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/roadmap-refresh.md +33 -31
- {roadmodel-0.2.94/docs/claude-commands → roadmodel-0.2.96/roadmodel/data}/roadmap-step.md +17 -8
- {roadmodel-0.2.94/docs/claude-commands → roadmodel-0.2.96/roadmodel/data}/roadmodel-upgrade.md +19 -1
- {roadmodel-0.2.94 → roadmodel-0.2.96}/scripts/update_projects.py +291 -1
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/__init__.py +1 -1
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/cli.py +37 -12
- roadmodel-0.2.96/src/roadmodel/ladder_layout.py +422 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/scoring.py +44 -4
- roadmodel-0.2.94/src/roadmodel/ladder_layout.py +0 -242
- {roadmodel-0.2.94 → roadmodel-0.2.96}/.github/README.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/.gitignore +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/LICENSE +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/NOTICE +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/README.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/benchmarks.json +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/claude-commands/roadmap-phase.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/claude-commands/roadmap-project.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/context-windows.json +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/effort-tokens.json +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/model-selector.txt +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/model-tier-cost-scale.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/settings-display.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/docs/user-context.example.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/hatch_build.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/infra/README.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/infra/gcp-killswitch/README.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/infra/supabase/README.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/benchmarks.json +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/context-windows.json +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/effort-tokens.json +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/model-selector.txt +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/model-tier-cost-scale.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/roadmap-phase.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/roadmap-project.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/settings-display.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/roadmodel/data/user-context.example.md +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/scripts/agent_context.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/scripts/agent_rules.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/scripts/updater_lock.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/scripts/upgrade_fleet.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/scripts/worktree_memory.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/__main__.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/config.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/cost.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/errors.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/mcp_server.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/private_files.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/__init__.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/_safety.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/anthropic.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/google.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/openai.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/openai_compatible.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/providers/registry.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/recommend.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/series.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/usage.py +0 -0
- {roadmodel-0.2.94 → roadmodel-0.2.96}/src/roadmodel/user_context.py +0 -0
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.5
|
|
2
2
|
Name: roadmodel
|
|
3
|
-
Version: 0.2.
|
|
3
|
+
Version: 0.2.96
|
|
4
4
|
Summary: BYO-key CLI that recommends the right AI model, platform, and settings for a prompt.
|
|
5
5
|
Project-URL: Homepage, https://roadmodel.ai
|
|
6
6
|
Project-URL: Repository, https://github.com/nathanramoscfa/roadmodel
|
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
{
|
|
2
2
|
"schema_version": "4",
|
|
3
3
|
"output_contract_version": 2,
|
|
4
|
-
"generated_at_utc": "2026-10-
|
|
4
|
+
"generated_at_utc": "2026-10-25T19:01:45Z",
|
|
5
5
|
"source_doc_sha256": {
|
|
6
6
|
"model-selector.txt": "e2ef6c5a18f05ff952fad9f576720b8fc4637f563f60ce5c077c5501fc21d1b7",
|
|
7
7
|
"model-tier-cost-scale.md": "dad707e59d0f3e75cf699d81368ba66783cc2a6a0bf625c430e8f884ef844a1c"
|
|
@@ -176,7 +176,7 @@ complexity, novel or not; its Model rationale names them), then:
|
|
|
176
176
|
for that classification (MCP tool: `roadmap_settings`), pasted as
|
|
177
177
|
printed with its **Runs:** line: Cost, Balanced and Quality, each
|
|
178
178
|
ranked three deep (Pick, Runner-up, and a Backup from another maker),
|
|
179
|
-
each option
|
|
179
|
+
each option with its own level, AA Index and price, read with every usage
|
|
180
180
|
pool at headroom, so the step's runner picks the column from the pools
|
|
181
181
|
on the day. A single-column table becomes the ladder: that is a
|
|
182
182
|
difference. Its models are always the latest version of their series.
|
|
@@ -203,31 +203,34 @@ complexity, novel or not; its Model rationale names them), then:
|
|
|
203
203
|
sentence, or its entry in the Model selection blocks) — a layout
|
|
204
204
|
fix, not a new pick, so it takes no `Settings updated` line.
|
|
205
205
|
- Every ladder, a step's Settings and a phase's Phase roadmap settings
|
|
206
|
-
alike, takes the
|
|
207
|
-
ladder` prints it:
|
|
208
|
-
**
|
|
209
|
-
|
|
210
|
-
|
|
211
|
-
|
|
212
|
-
|
|
213
|
-
|
|
214
|
-
|
|
215
|
-
`roadmodel relayout <every
|
|
216
|
-
|
|
217
|
-
table with Cost / Balanced / Quality columns
|
|
218
|
-
|
|
219
|
-
|
|
220
|
-
steps
|
|
221
|
-
|
|
222
|
-
|
|
223
|
-
|
|
224
|
-
|
|
225
|
-
|
|
226
|
-
|
|
227
|
-
|
|
228
|
-
|
|
229
|
-
|
|
230
|
-
|
|
206
|
+
alike, takes the table layout (roadmodel 0.2.95 on), as `roadmodel
|
|
207
|
+
ladder` prints it: Cost, Balanced and Quality columns (the one its
|
|
208
|
+
**Runs:** line names marked ✓), each rank a bold group row (**Pick**,
|
|
209
|
+
**Runner‑up**, **Backup**) over indented Model / Platform / Level / AA /
|
|
210
|
+
Price rows — Level the bare level of the platform's own dial, AA its AA
|
|
211
|
+
Intelligence Index there (`not measured` where AA has no row), Price its
|
|
212
|
+
blended list price per 1M tokens times the tokens it uses at that level —
|
|
213
|
+
then a note line, a `**Conversation:**` line and the **Runs:** line. It
|
|
214
|
+
has no Thinking row: Thinking is always On where the platform has the
|
|
215
|
+
toggle. Before re-selecting anything, run `roadmodel relayout <every
|
|
216
|
+
roadmap file>`: it rewrites each ladder roadmodel 0.2.86–0.2.94 wrote (a
|
|
217
|
+
`| Rank |` or `| Setting |` table with Cost / Balanced / Quality columns,
|
|
218
|
+
or 0.2.94's `- Pick:` lists) as that table, every model, platform, level
|
|
219
|
+
and AA Index kept as written and Price filled in from the catalog. It
|
|
220
|
+
reaches only steps that read `Not started` and phases whose roadmap does
|
|
221
|
+
not exist: a Complete step's ladder is the record of what ran, an `In
|
|
222
|
+
progress` step's is what an agent is running, and a written phase's
|
|
223
|
+
records what wrote its roadmap, so those keep their layout. That is a
|
|
224
|
+
layout fix, not a new pick, and takes no `Settings updated` line. A
|
|
225
|
+
ladder written before roadmodel 0.2.93 has Model / Backup / Platform /
|
|
226
|
+
Effort (/ AA Index) / Thinking rows and no Runner-up: on a step that has
|
|
227
|
+
not shipped, or a phase not yet written, replace it with the ladder
|
|
228
|
+
printed for that classification, even where every pick still matches —
|
|
229
|
+
a layout fix, not a new pick, so it takes no `Settings updated` line
|
|
230
|
+
unless the Pick of the column its **Runs:** line names changed. Without
|
|
231
|
+
roadmodel, leave every ladder as it is.
|
|
232
|
+
- A Complete step's ladder keeps its values and its layout (see below),
|
|
233
|
+
with its `> Ran` and `> Settings updated` lines beneath it.
|
|
231
234
|
- Keep the step's selection block in the Model selection blocks equal
|
|
232
235
|
to the column its **Runs:** line names (its `BACKUP:` line to that
|
|
233
236
|
column's Backup), and its Summary Table row's Model to that column's
|
|
@@ -242,9 +245,8 @@ complexity, novel or not; its Model rationale names them), then:
|
|
|
242
245
|
included, with no `Settings updated` line.
|
|
243
246
|
- If it matches, leave the step exactly as it is.
|
|
244
247
|
- **Never touch a step that reads `Complete`.** Its settings are the
|
|
245
|
-
record of what ran. Only the
|
|
246
|
-
|
|
247
|
-
value.
|
|
248
|
+
record of what ran. Only the Summary Table's column fix reaches it,
|
|
249
|
+
because it changes no value.
|
|
248
250
|
- Keep each step's Conversation row as written — it is session hygiene,
|
|
249
251
|
not a model choice.
|
|
250
252
|
|
|
@@ -279,10 +281,10 @@ sentences on why.
|
|
|
279
281
|
beneath it.
|
|
280
282
|
- A table that matches ⇒ leave the phase exactly as it is, except
|
|
281
283
|
that a ladder written before roadmodel 0.2.93 becomes the ranked
|
|
282
|
-
ladder, as a step's does, and
|
|
284
|
+
ladder, as a step's does, and its ladder takes the table layout.
|
|
283
285
|
- **Never add, re-pick or rewrite the table of a phase whose roadmap
|
|
284
286
|
exists.** Its table, or the lack of one, is the record of what wrote
|
|
285
|
-
that roadmap
|
|
287
|
+
that roadmap, layout included.
|
|
286
288
|
|
|
287
289
|
Print the phase · was · now · changed table.
|
|
288
290
|
|
|
@@ -101,16 +101,25 @@ Print one line: `Step {{STEP}} requires: Model <M> · Platform <P> ·
|
|
|
101
101
|
<the dials the table carries, e.g. Effort <E>[ · Thinking <T>], or
|
|
102
102
|
Intelligence <I>>. This session: <your own model> on <this surface>.`
|
|
103
103
|
|
|
104
|
+
**Thinking is always On** wherever the platform has a Thinking toggle
|
|
105
|
+
(Antigravity, the Gemini app, the APIs that expose one). A ladder carries
|
|
106
|
+
no Thinking row because it never recommends Off: every AA Index and Price
|
|
107
|
+
in it assumes thinking on. On such a platform, print `Thinking On` among
|
|
108
|
+
the dials and make sure it is set.
|
|
109
|
+
|
|
104
110
|
- **Ladders** (Cost / Balanced / Quality columns, roadmodel 0.2.86
|
|
105
111
|
on) carry every column the step may run, so pick one before the checks
|
|
106
|
-
below. From roadmodel 0.2.
|
|
107
|
-
|
|
108
|
-
|
|
109
|
-
|
|
110
|
-
|
|
111
|
-
|
|
112
|
-
|
|
113
|
-
|
|
112
|
+
below. From roadmodel 0.2.95 a ladder is one table: Cost, Balanced and
|
|
113
|
+
Quality are its columns, and each rank is a bold group row (**Pick**,
|
|
114
|
+
**Runner‑up**, **Backup**) over indented Model, Platform, Level, AA and
|
|
115
|
+
Price rows. Level is the bare level of that platform's own dial (`High`
|
|
116
|
+
is Intelligence High on Codex, Effort High on Claude Code and
|
|
117
|
+
Antigravity). The ✓ in a column's header marks the one its **Runs:**
|
|
118
|
+
line names. A 0.2.94 ladder is three lists instead (`- Pick:`,
|
|
119
|
+
`- Runner-up:`, `- Backup:` under each column's bold heading), and a
|
|
120
|
+
0.2.93 one a table with rows `1 · Pick`, `2 · Runner-up` and
|
|
121
|
+
`3 · Backup`, each followed by Platform, Effort, Thinking and AA Index
|
|
122
|
+
rows. An older ladder has Model /
|
|
114
123
|
Backup / Platform / Effort rows instead: its Model, Platform and Effort
|
|
115
124
|
cells are the Pick, its Backup the Backup (split into `Backup`, `Backup
|
|
116
125
|
on` and `Backup effort` rows on a Complete step since 0.2.93), and it
|
{roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/claude-commands}/roadmodel-upgrade.md
RENAMED
|
@@ -1,5 +1,5 @@
|
|
|
1
1
|
---
|
|
2
|
-
description: Upgrade roadmodel and shared agent context across registered projects and configured machines (usage: /roadmodel-upgrade [project dirs to register…] [--sync-only | --check] [--local-only] [--skip-roadmap-refresh] [--refresh-roadmaps] [--install-schedule HH:MM])
|
|
2
|
+
description: Upgrade roadmodel and shared agent context across registered projects and configured machines (usage: /roadmodel-upgrade [project dirs to register…] [--sync-only | --check] [--local-only] [--skip-roadmap-refresh] [--refresh-roadmaps] [--install-schedule HH:MM] [--install-sync-schedule [MINUTES]])
|
|
3
3
|
---
|
|
4
4
|
Upgrade roadmodel in every registered project's own environment, refresh its
|
|
5
5
|
`planning/` kit and agent commands, and maintain live shared rules, full
|
|
@@ -172,6 +172,24 @@ and registries are local; installing one schedule does not install another
|
|
|
172
172
|
on a peer. When the operator has requested automatic upkeep, install and
|
|
173
173
|
verify it within that authorization.
|
|
174
174
|
|
|
175
|
+
**Checkout sync (roadmodel 0.2.96 on).** A roadmap change merged on GitHub
|
|
176
|
+
reaches a project's checkout only when it pulls, so the editor kept showing
|
|
177
|
+
the old version. `--install-sync-schedule [MINUTES]` (default 15) registers
|
|
178
|
+
`--sync-checkouts` on its own schedule: launchd, Task Scheduler while the user
|
|
179
|
+
is logged on, or cron. Every full run also syncs, after its roadmap refresh.
|
|
180
|
+
For each registered checkout that sits on its default branch, it fetches
|
|
181
|
+
origin and fast-forwards (`git merge --ff-only`). It never merges, rebases,
|
|
182
|
+
stashes, resets or cleans. It leaves alone, and reports in the log, any
|
|
183
|
+
checkout that:
|
|
184
|
+
- is on a feature branch (an agent's step, which gets main when it rebases);
|
|
185
|
+
- has local commits;
|
|
186
|
+
- has a merge or rebase in progress;
|
|
187
|
+
- has uncommitted changes to a file the update changes.
|
|
188
|
+
|
|
189
|
+
Untracked files are never touched. `--uninstall-sync-schedule` removes it.
|
|
190
|
+
The operator asked for this on 2026-10-25: install it on every machine, peers
|
|
191
|
+
included (over SSH: `<python> <launcher> --install-sync-schedule`).
|
|
192
|
+
|
|
175
193
|
## 6. Refresh every project's roadmaps
|
|
176
194
|
|
|
177
195
|
`--refresh-roadmaps` runs `/roadmap-refresh` in every registered project on
|
{roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/templates}/phase-roadmap-template.md
RENAMED
|
@@ -757,24 +757,27 @@ that owns it.}}
|
|
|
757
757
|
**Runs:** line exactly as `roadmodel ladder --category
|
|
758
758
|
<c> --complexity <x> [--novel]` prints them for the
|
|
759
759
|
step's classification (MCP tool: `roadmap_settings`).
|
|
760
|
-
Layout (roadmodel 0.2.
|
|
761
|
-
|
|
762
|
-
|
|
763
|
-
|
|
764
|
-
|
|
765
|
-
platform
|
|
766
|
-
on Codex
|
|
767
|
-
|
|
768
|
-
|
|
769
|
-
|
|
770
|
-
|
|
771
|
-
|
|
772
|
-
|
|
773
|
-
|
|
774
|
-
|
|
775
|
-
|
|
776
|
-
|
|
777
|
-
|
|
760
|
+
Layout (roadmodel 0.2.95 on): one table whose columns
|
|
761
|
+
are Cost, Balanced and Quality (the column the **Runs:**
|
|
762
|
+
line names marked ✓). Each rank is a bold group row —
|
|
763
|
+
**Pick**, **Runner‑up**, **Backup** — over indented
|
|
764
|
+
field rows: Model (bold); Platform; Level, the bare
|
|
765
|
+
level the platform's own dial sets (`High` is
|
|
766
|
+
Intelligence High on Codex and Effort High on Claude
|
|
767
|
+
Code and Antigravity; `—` where the surface has no
|
|
768
|
+
dial); AA, the model's Artificial Analysis Intelligence
|
|
769
|
+
Index at that level (`not measured` where AA has no row
|
|
770
|
+
for it there); and Price, its blended list price per 1M
|
|
771
|
+
tokens (3 in : 1 out) times the tokens it uses at that
|
|
772
|
+
level (`*` where that use is not yet measured). There is
|
|
773
|
+
no Thinking row: Thinking is always On wherever the
|
|
774
|
+
platform has the toggle, and every AA Index and Price
|
|
775
|
+
already assumes it. A note line under the table explains
|
|
776
|
+
✓ and Price, and a `**Conversation:**` line closes the
|
|
777
|
+
ladder. The ranks follow the column's own rule, not the
|
|
778
|
+
AA Index: Cost ranks the cheapest first, so its AA Index
|
|
779
|
+
may rise down the column. The Backup reads `none` when
|
|
780
|
+
the Runner-up already holds the
|
|
778
781
|
only other maker's models. Every model is the latest
|
|
779
782
|
version of its series, never an older one (GPT-6 Sol
|
|
780
783
|
where GPT-6.1 Sol runs, Opus 5 where Opus 5.5 does).
|
|
@@ -875,29 +878,35 @@ that owns it.}}
|
|
|
875
878
|
-->
|
|
876
879
|
|
|
877
880
|
Settings — ladder ({{the default wherever roadmodel is
|
|
878
|
-
installed: the
|
|
881
|
+
installed: the table and **Runs:** line `roadmodel ladder
|
|
879
882
|
--category <c> --complexity <x> [--novel]` prints for the
|
|
880
883
|
step's classification, pasted as printed. The three table
|
|
881
884
|
shapes after it are the fallback where roadmodel cannot be
|
|
882
885
|
reached}}):
|
|
883
886
|
|
|
884
|
-
|
|
885
|
-
|
|
886
|
-
|
|
887
|
-
|
|
888
|
-
|
|
889
|
-
|
|
890
|
-
|
|
891
|
-
|
|
892
|
-
|
|
893
|
-
|
|
894
|
-
|
|
895
|
-
|
|
896
|
-
|
|
897
|
-
|
|
898
|
-
|
|
899
|
-
|
|
900
|
-
|
|
887
|
+
| | Cost | Balanced ✓ | Quality |
|
|
888
|
+
| -------------------------------- | --------------------- | --------------------- | --------------------- |
|
|
889
|
+
| **Pick** | | | |
|
|
890
|
+
| Model | **{{Model}}** | **{{Model}}** | **{{Model}}** |
|
|
891
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
892
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
893
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
894
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
895
|
+
| **Runner‑up** | | | |
|
|
896
|
+
| Model | **{{Model}}** | **{{Model}}** | **{{Model}}** |
|
|
897
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
898
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
899
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
900
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
901
|
+
| **Backup** | | | |
|
|
902
|
+
| Model | **{{Model or none}}** | **{{Model or none}}** | **{{Model or none}}** |
|
|
903
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
904
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
905
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
906
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
907
|
+
|
|
908
|
+
✓ runs by default. Price: list $ per 1M tokens (3 in : 1 out) × the tokens the
|
|
909
|
+
model uses at that level.
|
|
901
910
|
|
|
902
911
|
**Conversation:** New
|
|
903
912
|
|
|
@@ -1317,23 +1326,29 @@ printed with its **Runs:** line. Where roadmodel cannot be
|
|
|
1317
1326
|
reached, the single-column variant that matches PLATFORM,
|
|
1318
1327
|
carrying ONLY the rows that platform's panel exposes.}}
|
|
1319
1328
|
|
|
1320
|
-
|
|
1321
|
-
|
|
1322
|
-
|
|
1323
|
-
|
|
1324
|
-
|
|
1325
|
-
|
|
1326
|
-
|
|
1327
|
-
|
|
1328
|
-
|
|
1329
|
-
|
|
1330
|
-
|
|
1331
|
-
|
|
1332
|
-
|
|
1333
|
-
|
|
1334
|
-
|
|
1335
|
-
|
|
1336
|
-
|
|
1329
|
+
| | Cost | Balanced ✓ | Quality |
|
|
1330
|
+
| -------------------------------- | --------------------- | --------------------- | --------------------- |
|
|
1331
|
+
| **Pick** | | | |
|
|
1332
|
+
| Model | **{{Model}}** | **{{Model}}** | **{{Model}}** |
|
|
1333
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
1334
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
1335
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
1336
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
1337
|
+
| **Runner‑up** | | | |
|
|
1338
|
+
| Model | **{{Model}}** | **{{Model}}** | **{{Model}}** |
|
|
1339
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
1340
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
1341
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
1342
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
1343
|
+
| **Backup** | | | |
|
|
1344
|
+
| Model | **{{Model or none}}** | **{{Model or none}}** | **{{Model or none}}** |
|
|
1345
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
1346
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
1347
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
1348
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
1349
|
+
|
|
1350
|
+
✓ runs by default. Price: list $ per 1M tokens (3 in : 1 out) × the tokens the
|
|
1351
|
+
model uses at that level.
|
|
1337
1352
|
|
|
1338
1353
|
**Conversation:** New
|
|
1339
1354
|
|
|
@@ -1464,23 +1479,29 @@ this step's classification, pasted as printed with its
|
|
|
1464
1479
|
**Runs:** line; where roadmodel cannot be reached, the
|
|
1465
1480
|
single-column variant that matches PLATFORM.}}
|
|
1466
1481
|
|
|
1467
|
-
|
|
1468
|
-
|
|
1469
|
-
|
|
1470
|
-
|
|
1471
|
-
|
|
1472
|
-
|
|
1473
|
-
|
|
1474
|
-
|
|
1475
|
-
|
|
1476
|
-
|
|
1477
|
-
|
|
1478
|
-
|
|
1479
|
-
|
|
1480
|
-
|
|
1481
|
-
|
|
1482
|
-
|
|
1483
|
-
|
|
1482
|
+
| | Cost | Balanced ✓ | Quality |
|
|
1483
|
+
| -------------------------------- | --------------------- | --------------------- | --------------------- |
|
|
1484
|
+
| **Pick** | | | |
|
|
1485
|
+
| Model | **{{Model}}** | **{{Model}}** | **{{Model}}** |
|
|
1486
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
1487
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
1488
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
1489
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
1490
|
+
| **Runner‑up** | | | |
|
|
1491
|
+
| Model | **{{Model}}** | **{{Model}}** | **{{Model}}** |
|
|
1492
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
1493
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
1494
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
1495
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
1496
|
+
| **Backup** | | | |
|
|
1497
|
+
| Model | **{{Model or none}}** | **{{Model or none}}** | **{{Model or none}}** |
|
|
1498
|
+
| Platform | {{Platform}} | {{Platform}} | {{Platform}} |
|
|
1499
|
+
| Level | {{Level}} | {{Level}} | {{Level}} |
|
|
1500
|
+
| AA | {{AA Index}} | {{AA Index}} | {{AA Index}} |
|
|
1501
|
+
| Price | {{$ per 1M}} | {{$ per 1M}} | {{$ per 1M}} |
|
|
1502
|
+
|
|
1503
|
+
✓ runs by default. Price: list $ per 1M tokens (3 in : 1 out) × the tokens the
|
|
1504
|
+
model uses at that level.
|
|
1484
1505
|
|
|
1485
1506
|
**Conversation:** New
|
|
1486
1507
|
|
{roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/templates}/planning-kit-how-to-use.md
RENAMED
|
@@ -32,8 +32,8 @@ out would add cost for no benefit.
|
|
|
32
32
|
the step (step 1 below), then paste the ladder and **Runs:** line
|
|
33
33
|
`roadmodel ladder --category <c> --complexity <x> [--novel]` prints: the
|
|
34
34
|
Cost / Balanced / Quality picks /recommend shows, each ranked three deep (the
|
|
35
|
-
Pick, the Runner-up, and a Backup from another maker), each option
|
|
36
|
-
|
|
35
|
+
Pick, the Runner-up, and a Backup from another maker), each option with its
|
|
36
|
+
level, AA Index and price. It is local code — no API
|
|
37
37
|
call, no engine — and it
|
|
38
38
|
reads every usage pool at headroom, so the ladder outlives a pool's state: when
|
|
39
39
|
a pool reads `tight` or `exhausted`, `/roadmap-step` runs another column
|
{roadmodel-0.2.94/roadmodel/data → roadmodel-0.2.96/docs/templates}/project-roadmap-template.md
RENAMED
|
@@ -45,12 +45,13 @@ STYLE RULES (the AI MUST follow)
|
|
|
45
45
|
in Claude Code, which `/roadmap-phase` sets by itself),
|
|
46
46
|
with thinking on where the model has a Thinking toggle
|
|
47
47
|
(Opus 5.5 and the other models that always think have
|
|
48
|
-
none
|
|
48
|
+
none, and the ladder carries no Thinking row: it is always On where a
|
|
49
|
+
toggle exists). Where `roadmodel` is installed, the table
|
|
49
50
|
is the ladder `roadmodel ladder --category planning
|
|
50
51
|
--complexity high --phase` prints, pasted as printed with
|
|
51
52
|
its **Runs:** line: Cost, Balanced and Quality, each ranked
|
|
52
53
|
three deep (Pick, Runner-up, and a Backup from another
|
|
53
|
-
maker), each option with its AA Index, so a pool that reads `tight` or
|
|
54
|
+
maker), each option with its level, AA Index and price, so a pool that reads `tight` or
|
|
54
55
|
`exhausted` on the day moves the session to another column
|
|
55
56
|
instead of the table being refreshed. It ranks on measured
|
|
56
57
|
benchmarks, so a stale catalog letter cannot steer it, and
|
|
@@ -249,23 +250,29 @@ dependency graph in §5 is honoured.
|
|
|
249
250
|
**Phase roadmap settings** — the session that writes
|
|
250
251
|
`phase01-roadmap.md` (`/roadmap-phase 1`):
|
|
251
252
|
|
|
252
|
-
|
|
253
|
-
|
|
254
|
-
|
|
255
|
-
|
|
256
|
-
|
|
257
|
-
|
|
258
|
-
|
|
259
|
-
|
|
260
|
-
|
|
261
|
-
|
|
262
|
-
|
|
263
|
-
|
|
264
|
-
|
|
265
|
-
|
|
266
|
-
|
|
267
|
-
|
|
268
|
-
|
|
253
|
+
| | Cost | Balanced | Quality ✓ |
|
|
254
|
+
| -------------------------------- | ------------------- | ------------------- | ------------------- |
|
|
255
|
+
| **Pick** | | | |
|
|
256
|
+
| Model | **<Model>** | **<Model>** | **<Model>** |
|
|
257
|
+
| Platform | <Platform> | <Platform> | <Platform> |
|
|
258
|
+
| Level | <Level> | <Level> | <Level> |
|
|
259
|
+
| AA | <AA Index> | <AA Index> | <AA Index> |
|
|
260
|
+
| Price | <$ per 1M> | <$ per 1M> | <$ per 1M> |
|
|
261
|
+
| **Runner‑up** | | | |
|
|
262
|
+
| Model | **<Model>** | **<Model>** | **<Model>** |
|
|
263
|
+
| Platform | <Platform> | <Platform> | <Platform> |
|
|
264
|
+
| Level | <Level> | <Level> | <Level> |
|
|
265
|
+
| AA | <AA Index> | <AA Index> | <AA Index> |
|
|
266
|
+
| Price | <$ per 1M> | <$ per 1M> | <$ per 1M> |
|
|
267
|
+
| **Backup** | | | |
|
|
268
|
+
| Model | **<Model or none>** | **<Model or none>** | **<Model or none>** |
|
|
269
|
+
| Platform | <Platform> | <Platform> | <Platform> |
|
|
270
|
+
| Level | <Level> | <Level> | <Level> |
|
|
271
|
+
| AA | <AA Index> | <AA Index> | <AA Index> |
|
|
272
|
+
| Price | <$ per 1M> | <$ per 1M> | <$ per 1M> |
|
|
273
|
+
|
|
274
|
+
✓ runs by default. Price: list $ per 1M tokens (3 in : 1 out) × the tokens the
|
|
275
|
+
model uses at that level.
|
|
269
276
|
|
|
270
277
|
**Conversation:** New
|
|
271
278
|
|
|
@@ -323,23 +330,29 @@ key pays for the platform.>
|
|
|
323
330
|
**Phase roadmap settings** — the session that writes
|
|
324
331
|
`phase02-roadmap.md` (`/roadmap-phase 2`):
|
|
325
332
|
|
|
326
|
-
|
|
327
|
-
|
|
328
|
-
|
|
329
|
-
|
|
330
|
-
|
|
331
|
-
|
|
332
|
-
|
|
333
|
-
|
|
334
|
-
|
|
335
|
-
|
|
336
|
-
|
|
337
|
-
|
|
338
|
-
|
|
339
|
-
|
|
340
|
-
|
|
341
|
-
|
|
342
|
-
|
|
333
|
+
| | Cost | Balanced | Quality ✓ |
|
|
334
|
+
| -------------------------------- | ----- | -------- | --------- |
|
|
335
|
+
| **Pick** | | | |
|
|
336
|
+
| Model | **…** | **…** | **…** |
|
|
337
|
+
| Platform | … | … | … |
|
|
338
|
+
| Level | … | … | … |
|
|
339
|
+
| AA | … | … | … |
|
|
340
|
+
| Price | … | … | … |
|
|
341
|
+
| **Runner‑up** | | | |
|
|
342
|
+
| Model | **…** | **…** | **…** |
|
|
343
|
+
| Platform | … | … | … |
|
|
344
|
+
| Level | … | … | … |
|
|
345
|
+
| AA | … | … | … |
|
|
346
|
+
| Price | … | … | … |
|
|
347
|
+
| **Backup** | | | |
|
|
348
|
+
| Model | **…** | **…** | **…** |
|
|
349
|
+
| Platform | … | … | … |
|
|
350
|
+
| Level | … | … | … |
|
|
351
|
+
| AA | … | … | … |
|
|
352
|
+
| Price | … | … | … |
|
|
353
|
+
|
|
354
|
+
✓ runs by default. Price: list $ per 1M tokens (3 in : 1 out) × the tokens the
|
|
355
|
+
model uses at that level.
|
|
343
356
|
|
|
344
357
|
**Conversation:** New
|
|
345
358
|
|
|
@@ -371,23 +384,29 @@ key pays for the platform.>
|
|
|
371
384
|
**Phase roadmap settings** — the session that writes
|
|
372
385
|
`phaseNN-roadmap.md` (`/roadmap-phase N`):
|
|
373
386
|
|
|
374
|
-
|
|
375
|
-
|
|
376
|
-
|
|
377
|
-
|
|
378
|
-
|
|
379
|
-
|
|
380
|
-
|
|
381
|
-
|
|
382
|
-
|
|
383
|
-
|
|
384
|
-
|
|
385
|
-
|
|
386
|
-
|
|
387
|
-
|
|
388
|
-
|
|
389
|
-
|
|
390
|
-
|
|
387
|
+
| | Cost | Balanced | Quality ✓ |
|
|
388
|
+
| -------------------------------- | ----- | -------- | --------- |
|
|
389
|
+
| **Pick** | | | |
|
|
390
|
+
| Model | **…** | **…** | **…** |
|
|
391
|
+
| Platform | … | … | … |
|
|
392
|
+
| Level | … | … | … |
|
|
393
|
+
| AA | … | … | … |
|
|
394
|
+
| Price | … | … | … |
|
|
395
|
+
| **Runner‑up** | | | |
|
|
396
|
+
| Model | **…** | **…** | **…** |
|
|
397
|
+
| Platform | … | … | … |
|
|
398
|
+
| Level | … | … | … |
|
|
399
|
+
| AA | … | … | … |
|
|
400
|
+
| Price | … | … | … |
|
|
401
|
+
| **Backup** | | | |
|
|
402
|
+
| Model | **…** | **…** | **…** |
|
|
403
|
+
| Platform | … | … | … |
|
|
404
|
+
| Level | … | … | … |
|
|
405
|
+
| AA | … | … | … |
|
|
406
|
+
| Price | … | … | … |
|
|
407
|
+
|
|
408
|
+
✓ runs by default. Price: list $ per 1M tokens (3 in : 1 out) × the tokens the
|
|
409
|
+
model uses at that level.
|
|
391
410
|
|
|
392
411
|
**Conversation:** New
|
|
393
412
|
|
|
@@ -28,8 +28,8 @@ Step 1 has `**Status:** Not started` directly under the `## Step 1`
|
|
|
28
28
|
heading. If Stage 3 is a bare "OPEN THE PR", Stage 6 tells you to put
|
|
29
29
|
findings in a "Follow-ups (non-blocking)" note after the completion
|
|
30
30
|
line, a step's Status line sits after its `**Deploys:**` line, its
|
|
31
|
-
Settings tables have no `Backup` row, or its
|
|
32
|
-
|
|
31
|
+
Settings tables have no `Backup` row, or its ladder tables have no
|
|
32
|
+
`Price` row, STOP
|
|
33
33
|
and tell me the kit is stale — do not write the roadmap from it.
|
|
34
34
|
|
|
35
35
|
Step 0b — roadmaps live in `docs/roadmap/`: `ROADMAP.md` and every
|
|
@@ -70,11 +70,11 @@ own model> on <this surface>.`
|
|
|
70
70
|
the model, print `unverified`, ask me which I picked, and go on with my
|
|
71
71
|
answer.
|
|
72
72
|
- A ladder (Cost / Balanced / Quality columns): each column ranks a
|
|
73
|
-
Pick, a Runner-up and a Backup
|
|
74
|
-
|
|
75
|
-
0.2.94
|
|
76
|
-
|
|
77
|
-
|
|
73
|
+
Pick, a Runner-up and a Backup — from roadmodel 0.2.95 a bold group
|
|
74
|
+
row each, over indented Model / Platform / Level / AA / Price rows
|
|
75
|
+
(0.2.94 wrote them as `- Pick:` lists, 0.2.93 as `1 · Pick` rows,
|
|
76
|
+
and older ladders Model / Backup rows with no Runner-up) — and its
|
|
77
|
+
**Runs:** line names the
|
|
78
78
|
column this session runs (Quality, for a phase roadmap). Apply that
|
|
79
79
|
line's rule to the Usage-pool status table in
|
|
80
80
|
`planning/user-context.md`: a pool that reads `tight` or `exhausted`
|
|
@@ -123,7 +123,7 @@ that classification (MCP tool: `roadmap_settings`), pasted as printed
|
|
|
123
123
|
with its **Runs:** line. It carries the full recommendation /recommend
|
|
124
124
|
shows: Cost, Balanced and Quality, each ranked three deep (the Pick,
|
|
125
125
|
the Runner-up, and a Backup from another maker), each option on one
|
|
126
|
-
|
|
126
|
+
its own level, AA Index and price. The step's runner picks the column from the
|
|
127
127
|
usage pools on the day, so the roadmap needs no refresh when a pool
|
|
128
128
|
tightens. Its models are always the latest version of their series.
|
|
129
129
|
The Model rationale explains the classification and what sets the
|