bigpowers 2.45.1 → 2.45.3

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.
package/.pi/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "bigpowers",
3
- "version": "2.45.1",
3
+ "version": "2.45.3",
4
4
  "description": "72 skills — 72 agent skills synthesizing 17 years of software engineering discipline into a prescriptive methodology for solo developers",
5
5
  "keywords": [
6
6
  "pi-package"
package/CHANGELOG.md CHANGED
@@ -1,3 +1,17 @@
1
+ ## [2.45.3](https://github.com/danielvm-git/bigpowers/compare/v2.45.2...v2.45.3) (2026-07-03)
2
+
3
+
4
+ ### Bug Fixes
5
+
6
+ * **sync:** sync generated CLAUDE.md and opencode.json artifacts ([f018e27](https://github.com/danielvm-git/bigpowers/commit/f018e27cf71d61190bf58c1ef884e100e15a8aee))
7
+
8
+ ## [2.45.2](https://github.com/danielvm-git/bigpowers/compare/v2.45.1...v2.45.2) (2026-07-03)
9
+
10
+
11
+ ### Bug Fixes
12
+
13
+ * **sync:** sync stale .pi/package.json version to 2.45.1 ([4901405](https://github.com/danielvm-git/bigpowers/commit/49014051a42a63c9499d8aedc379ae7348fbb235))
14
+
1
15
  ## [2.45.1](https://github.com/danielvm-git/bigpowers/compare/v2.45.0...v2.45.1) (2026-07-02)
2
16
 
3
17
 
package/CLAUDE.md CHANGED
@@ -64,3 +64,241 @@ Before any task, run this sequence — not optional:
64
64
  - Run tests after every change. Show evidence before declaring done.
65
65
  - One clarifying question beats a wrong assumption baked into 200 lines.
66
66
  - All written output (plans, specs, investigations) goes in specs/.
67
+
68
+ ## bts toolchain
69
+
70
+ `bts` is installed. Prefer its verbs over ad-hoc shell commands.
71
+
72
+ | Task | Command | Avoid |
73
+ |------|---------|-------|
74
+ | Search code | `bts find --print <pattern>` | grep / find / cat |
75
+ | Interactive search | `bts find <pattern>` | manual grep pipes |
76
+ | Compress for context | `bts compress <file>` or `cmd \| bts compress` | summarising by hand |
77
+ | Repo map | `bts map` | listing files by hand |
78
+ | Library docs | `bts docs <lib>` | guessing from training data |
79
+ | Package source | `bts src <pkg>` | git clone |
80
+ | Toolchain health | `bts doctor` | which / command -v |
81
+
82
+ **Rules**
83
+ - Search with `bts find` before opening files to locate a symbol or pattern.
84
+ - Pipe anything > 200 lines through `bts compress` before adding to context.
85
+ - Run `bts map` when asked for a repo overview.
86
+ - Use `bts docs <lib>` before answering questions about library APIs.
87
+ - If a tool is missing, say so and run `bts doctor` — do not silently substitute.
88
+
89
+ <!-- BEGIN sqz-claude-guidance (auto-installed by sqz init; remove this block to disable) -->
90
+
91
+ ## sqz — Context Compression (READ FIRST)
92
+
93
+ sqz is installed in this project. It compresses tool output so large
94
+ files, long logs, and verbose command output cost far fewer tokens.
95
+ There are **two ways** sqz is wired in, and you should prefer each
96
+ one in the situations below.
97
+
98
+ ### Preferred tools (MCP)
99
+
100
+ The `sqz-mcp` server is registered in this project's MCP config. It
101
+ exposes three read-only tools that compress their output through the
102
+ sqz pipeline:
103
+
104
+ - **`sqz_read_file`** — read a file from disk and return a compressed
105
+ view. **PREFER this over the built-in `Read` tool** for any file
106
+ larger than ~2KB or any file you might read more than once in the
107
+ same session. Repeat reads return a 13-token `§ref:HASH§` reference
108
+ instead of the full content.
109
+
110
+ - **`sqz_grep`** — search files for a literal string or regex.
111
+ **PREFER this over the built-in `Grep`** for anything that might
112
+ match more than a handful of lines. Caps at 200 matches by default;
113
+ raise with `max_matches` if needed.
114
+
115
+ - **`sqz_list_dir`** — list a directory. Skips `.git`, `node_modules`,
116
+ `target`, `dist`, `build`, `vendor`, `__pycache__` so the output
117
+ stays focused. **PREFER this over `ls -la` via Bash** when you want
118
+ to see a project layout.
119
+
120
+ The built-in `Read`, `Grep`, `Glob` tools remain available. Use them for:
121
+ - Tiny config files (<1KB) where compression can't help.
122
+ - Byte-exact reads you'll hash or diff (lockfiles, signatures).
123
+ - Globbing (sqz has no glob tool; `Glob` is still the right choice).
124
+
125
+ ### Bash commands (hooked automatically)
126
+
127
+ When you run a shell command through the `Bash` tool, a PreToolUse hook
128
+ rewrites it to pipe output through `sqz compress`. This is transparent:
129
+ you don't need to remember to add anything, but it's useful to know
130
+ that these commands get compressed automatically:
131
+
132
+ ```bash
133
+ git status # → git status 2>&1 | sqz compress --cmd git
134
+ cargo test # → cargo test 2>&1 | sqz compress --cmd cargo
135
+ docker ps # → docker ps 2>&1 | sqz compress --cmd docker
136
+ kubectl get pods # → kubectl get pods 2>&1 | sqz compress --cmd kubectl
137
+ ```
138
+
139
+ The rewrite is skipped for interactive commands (`vim`, `ssh`,
140
+ `python`), compound commands (`a && b`, `a > file.txt`), and anything
141
+ already going through sqz.
142
+
143
+ ### Escape hatch — when you see a `§ref:HASH§` token
144
+
145
+ If tool output contains a `§ref:a1b2c3d4§` token and you need the full
146
+ content it points at, resolve it. Three equivalent ways:
147
+
148
+ - Shell: `/Users/danielvm/.local/bin/sqz expand a1b2c3d4` (or paste the whole token
149
+ `/Users/danielvm/.local/bin/sqz expand §ref:a1b2c3d4§`).
150
+ - MCP tool: call `expand` with `{ "prefix": "a1b2c3d4" }`.
151
+ - To get uncompressed output for one command: prefix it with
152
+ `SQZ_NO_DEDUP=1` (e.g. `SQZ_NO_DEDUP=1 git log | sqz compress`).
153
+
154
+ If the compressed output is actively making the task harder (looping
155
+ on refs, small retries replacing one big read), call the `passthrough`
156
+ MCP tool to get raw text.
157
+
158
+ ### When NOT to use sqz tools
159
+
160
+ - Writing or editing files — use the built-in `Write`/`Edit` tools.
161
+ sqz has no write tools (by design; see issue #5 follow-up).
162
+ - Running commands interactively or in watch mode.
163
+ - Reading very small files (<1KB) where compression can't help.
164
+
165
+ <!-- END sqz-claude-guidance -->
166
+
167
+ <!-- rtk-instructions v2 -->
168
+ # RTK (Rust Token Killer) - Token-Optimized Commands
169
+
170
+ ## Golden Rule
171
+
172
+ **Always prefix commands with `rtk`**. If RTK has a dedicated filter, it uses it. If not, it passes through unchanged. This means RTK is always safe to use.
173
+
174
+ **Important**: Even in command chains with `&&`, use `rtk`:
175
+ ```bash
176
+ # ❌ Wrong
177
+ git add . && git commit -m "msg" && git push
178
+
179
+ # ✅ Correct
180
+ rtk git add . && rtk git commit -m "msg" && rtk git push
181
+ ```
182
+
183
+ ## RTK Commands by Workflow
184
+
185
+ ### Build & Compile (80-90% savings)
186
+ ```bash
187
+ rtk cargo build # Cargo build output
188
+ rtk cargo check # Cargo check output
189
+ rtk cargo clippy # Clippy warnings grouped by file (80%)
190
+ rtk tsc # TypeScript errors grouped by file/code (83%)
191
+ rtk lint # ESLint/Biome violations grouped (84%)
192
+ rtk prettier --check # Files needing format only (70%)
193
+ rtk next build # Next.js build with route metrics (87%)
194
+ ```
195
+
196
+ ### Test (60-99% savings)
197
+ ```bash
198
+ rtk cargo test # Cargo test failures only (90%)
199
+ rtk go test # Go test failures only (90%)
200
+ rtk jest # Jest failures only (99.5%)
201
+ rtk vitest # Vitest failures only (99.5%)
202
+ rtk playwright test # Playwright failures only (94%)
203
+ rtk pytest # Python test failures only (90%)
204
+ rtk rake test # Ruby test failures only (90%)
205
+ rtk rspec # RSpec test failures only (60%)
206
+ rtk test <cmd> # Generic test wrapper - failures only
207
+ ```
208
+
209
+ ### Git (59-80% savings)
210
+ ```bash
211
+ rtk git status # Compact status
212
+ rtk git log # Compact log (works with all git flags)
213
+ rtk git diff # Compact diff (80%)
214
+ rtk git show # Compact show (80%)
215
+ rtk git add # Ultra-compact confirmations (59%)
216
+ rtk git commit # Ultra-compact confirmations (59%)
217
+ rtk git push # Ultra-compact confirmations
218
+ rtk git pull # Ultra-compact confirmations
219
+ rtk git branch # Compact branch list
220
+ rtk git fetch # Compact fetch
221
+ rtk git stash # Compact stash
222
+ rtk git worktree # Compact worktree
223
+ ```
224
+
225
+ Note: Git passthrough works for ALL subcommands, even those not explicitly listed.
226
+
227
+ ### GitHub (26-87% savings)
228
+ ```bash
229
+ rtk gh pr view <num> # Compact PR view (87%)
230
+ rtk gh pr checks # Compact PR checks (79%)
231
+ rtk gh run list # Compact workflow runs (82%)
232
+ rtk gh issue list # Compact issue list (80%)
233
+ rtk gh api # Compact API responses (26%)
234
+ ```
235
+
236
+ ### JavaScript/TypeScript Tooling (70-90% savings)
237
+ ```bash
238
+ rtk pnpm list # Compact dependency tree (70%)
239
+ rtk pnpm outdated # Compact outdated packages (80%)
240
+ rtk pnpm install # Compact install output (90%)
241
+ rtk npm run <script> # Compact npm script output
242
+ rtk npx <cmd> # Compact npx command output
243
+ rtk prisma # Prisma without ASCII art (88%)
244
+ ```
245
+
246
+ ### Files & Search (60-75% savings)
247
+ ```bash
248
+ rtk ls <path> # Tree format, compact (65%)
249
+ rtk read <file> # Code reading with filtering (60%)
250
+ rtk grep <pattern> # Search grouped by file (75%). Format flags (-c, -l, -L, -o, -Z) run raw.
251
+ rtk find <pattern> # Find grouped by directory (70%)
252
+ ```
253
+
254
+ ### Analysis & Debug (70-90% savings)
255
+ ```bash
256
+ rtk err <cmd> # Filter errors only from any command
257
+ rtk log <file> # Deduplicated logs with counts
258
+ rtk json <file> # JSON structure without values
259
+ rtk deps # Dependency overview
260
+ rtk env # Environment variables compact
261
+ rtk summary <cmd> # Smart summary of command output
262
+ rtk diff # Ultra-compact diffs
263
+ ```
264
+
265
+ ### Infrastructure (85% savings)
266
+ ```bash
267
+ rtk docker ps # Compact container list
268
+ rtk docker images # Compact image list
269
+ rtk docker logs <c> # Deduplicated logs
270
+ rtk kubectl get # Compact resource list
271
+ rtk kubectl logs # Deduplicated pod logs
272
+ ```
273
+
274
+ ### Network (65-70% savings)
275
+ ```bash
276
+ rtk curl <url> # Compact HTTP responses (70%)
277
+ rtk wget <url> # Compact download output (65%)
278
+ ```
279
+
280
+ ### Meta Commands
281
+ ```bash
282
+ rtk gain # View token savings statistics
283
+ rtk gain --history # View command history with savings
284
+ rtk discover # Analyze Claude Code sessions for missed RTK usage
285
+ rtk proxy <cmd> # Run command without filtering (for debugging)
286
+ rtk init # Add RTK instructions to CLAUDE.md
287
+ rtk init --global # Add RTK to ~/.claude/CLAUDE.md
288
+ ```
289
+
290
+ ## Token Savings Overview
291
+
292
+ | Category | Commands | Typical Savings |
293
+ |----------|----------|-----------------|
294
+ | Tests | vitest, playwright, cargo test | 90-99% |
295
+ | Build | next, tsc, lint, prettier | 70-87% |
296
+ | Git | status, log, diff, add, commit | 59-80% |
297
+ | GitHub | gh pr, gh run, gh issue | 26-87% |
298
+ | Package Managers | pnpm, npm, npx | 70-90% |
299
+ | Files | ls, read, grep, find | 60-75% |
300
+ | Infrastructure | docker, kubectl | 85% |
301
+ | Network | curl, wget | 65-70% |
302
+
303
+ Overall average: **60-90% token reduction** on common development operations.
304
+ <!-- /rtk-instructions -->
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "bigpowers",
3
- "version": "2.45.1",
3
+ "version": "2.45.3",
4
4
  "description": "72 agent skills synthesizing 17 years of software engineering discipline into a prescriptive methodology for solo developers",
5
5
  "main": "index.js",
6
6
  "scripts": {