warp-os 1.1.0

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.
Files changed (49) hide show
  1. package/CHANGELOG.md +327 -0
  2. package/LICENSE +21 -0
  3. package/README.md +308 -0
  4. package/VERSION +1 -0
  5. package/agents/warp-browse.md +715 -0
  6. package/agents/warp-build-code.md +1299 -0
  7. package/agents/warp-orchestrator.md +515 -0
  8. package/agents/warp-plan-architect.md +929 -0
  9. package/agents/warp-plan-brainstorm.md +876 -0
  10. package/agents/warp-plan-design.md +1458 -0
  11. package/agents/warp-plan-onboarding.md +732 -0
  12. package/agents/warp-plan-optimize-adversarial.md +81 -0
  13. package/agents/warp-plan-optimize.md +354 -0
  14. package/agents/warp-plan-scope.md +806 -0
  15. package/agents/warp-plan-security.md +1274 -0
  16. package/agents/warp-plan-testdesign.md +1228 -0
  17. package/agents/warp-qa-debug-adversarial.md +90 -0
  18. package/agents/warp-qa-debug.md +793 -0
  19. package/agents/warp-qa-test-adversarial.md +89 -0
  20. package/agents/warp-qa-test.md +1054 -0
  21. package/agents/warp-release-update.md +1189 -0
  22. package/agents/warp-setup.md +1216 -0
  23. package/agents/warp-upgrade.md +334 -0
  24. package/bin/cli.js +44 -0
  25. package/bin/hooks/_warp_html.sh +291 -0
  26. package/bin/hooks/_warp_json.sh +67 -0
  27. package/bin/hooks/consistency-check.sh +92 -0
  28. package/bin/hooks/identity-briefing.sh +89 -0
  29. package/bin/hooks/identity-foundation.sh +37 -0
  30. package/bin/install.js +343 -0
  31. package/dist/warp-browse/SKILL.md +727 -0
  32. package/dist/warp-build-code/SKILL.md +1316 -0
  33. package/dist/warp-orchestrator/SKILL.md +527 -0
  34. package/dist/warp-plan-architect/SKILL.md +943 -0
  35. package/dist/warp-plan-brainstorm/SKILL.md +890 -0
  36. package/dist/warp-plan-design/SKILL.md +1473 -0
  37. package/dist/warp-plan-onboarding/SKILL.md +742 -0
  38. package/dist/warp-plan-optimize/SKILL.md +364 -0
  39. package/dist/warp-plan-scope/SKILL.md +820 -0
  40. package/dist/warp-plan-security/SKILL.md +1286 -0
  41. package/dist/warp-plan-testdesign/SKILL.md +1244 -0
  42. package/dist/warp-qa-debug/SKILL.md +805 -0
  43. package/dist/warp-qa-test/SKILL.md +1070 -0
  44. package/dist/warp-release-update/SKILL.md +1211 -0
  45. package/dist/warp-setup/SKILL.md +1229 -0
  46. package/dist/warp-upgrade/SKILL.md +345 -0
  47. package/package.json +40 -0
  48. package/shared/project-hooks.json +32 -0
  49. package/shared/tier1-engineering-constitution.md +176 -0
@@ -0,0 +1,334 @@
1
+ ---
2
+ name: warp-upgrade
3
+ description: >-
4
+ One-command Warp upgrade from inside Claude Code. Pulls latest, rebuilds, reinstalls, shows changelog, detects migration requirements, and offers to re-run /warp-setup on the current project. No terminal switching needed.
5
+ ---
6
+
7
+ <!-- ═══════════════════════════════════════════════════════════ -->
8
+ <!-- TIER 1 — Engineering Foundation. Generated by build.sh -->
9
+ <!-- ═══════════════════════════════════════════════════════════ -->
10
+
11
+
12
+ # Warp Engineering Foundation
13
+
14
+ Universal principles for every agent in the Warp pipeline. Tier 1: highest authority.
15
+
16
+ ---
17
+
18
+ ## Core Principles
19
+
20
+ **Clarity over cleverness.** Optimize for "I can understand this in six months."
21
+
22
+ **Explicit contracts between layers.** Modules communicate through defined interfaces. Swap persistence without touching the service layer.
23
+
24
+ **Every component earns its place.** No speculative code. If a feature isn't in the current or next phase, it doesn't exist in code.
25
+
26
+ **Fail loud, recover gracefully.** Never swallow errors silently. User-facing experience degrades gracefully — stale-data indicator, not a crash.
27
+
28
+ **Prefer reversible decisions.** When two approaches are equivalent, choose the one that can be undone.
29
+
30
+ **Security is structural.** Designed for the most restrictive phase, enforced from the earliest.
31
+
32
+ **AI is a tool, not an authority.** AI agents accelerate development but do not make architectural decisions autonomously. Every significant design decision is reviewed by the user before it ships.
33
+
34
+ ---
35
+
36
+ ## Bias Classification
37
+
38
+ When the same AI system writes code, writes tests, and evaluates its own output, shared biases create blind spots.
39
+
40
+ | Level | Definition | Trust |
41
+ |-------|-----------|-------|
42
+ | **L1** | Deterministic. Binary pass/fail. Zero AI judgment. | Highest |
43
+ | **L2** | AI interpretation anchored to verifiable external source. | Medium |
44
+ | **L3** | AI evaluating AI. Both sides share training biases. | Lowest |
45
+
46
+ **L1 Imperative:** Every quality gate that CAN be L1 MUST be L1. L3 is the outer layer, never the only layer. When L1 is unavailable, use L2 (grounded in external docs). Fall back to L3 only when no external anchor exists.
47
+
48
+ ---
49
+
50
+ ## Completeness
51
+
52
+ AI compresses implementation 10-100x. Always choose the complete option. Full coverage, hardened behavior, robust edge cases. The delta between "good enough" and "complete" is minutes, not days.
53
+
54
+ Never recommend the less-complete option. Never skip edge cases. Never defer what can be done now.
55
+
56
+ ---
57
+
58
+ ## Quality Gates
59
+
60
+ **Hard Gate** — blocks progression. Between major phases. Present output, ask the user: A) Approve, B) Revise, C) Restart. MUST get user input.
61
+
62
+ **Soft Gate** — warns but allows. Between minor steps. Proceed if quality criteria met; warn and get input if not.
63
+
64
+ **Completeness Gate** — final check before artifact write. Verify no empty sections, key decisions explicit. Fix before writing.
65
+
66
+ ---
67
+
68
+ ## Escalation
69
+
70
+ Always OK to stop and escalate. Bad work is worse than no work.
71
+
72
+ **STOP if:** 3 failed attempts at the same problem, uncertain about security-sensitive changes, scope exceeds what you can verify, or a decision requires domain knowledge you don't have.
73
+
74
+ ---
75
+
76
+ ## External Data Gate
77
+
78
+ When a task requires real-world data or domain knowledge that cannot be derived from code, docs, or git history — PAUSE and ask the user. Never hallucinate fixtures or APIs. Check docs via Context7 or saved files before writing code that touches external services.
79
+
80
+ ---
81
+
82
+ ## Error Severity
83
+
84
+ | Tier | Definition | Response |
85
+ |------|-----------|----------|
86
+ | T1 | Normal variance (cache miss, retry succeeded) | Log, no action |
87
+ | T2 | Degraded capability (stale data served, fallback active) | Log, degrade visibly |
88
+ | T3 | Operation failed (invalid input, auth rejected) | Log, return error, continue |
89
+ | T4 | Subsystem non-functional (DB unreachable, corrupt state) | Log, halt subsystem, alert |
90
+
91
+ ---
92
+
93
+ ## Universal Engineering Principles
94
+
95
+ - Assert outcomes, not implementation. Test "input produces output" — not "function X calls Y."
96
+ - Each test is independent. No shared state or execution order dependencies.
97
+ - Mock at the system boundary, not internal helpers.
98
+ - Expected values are hardcoded from the spec, never recalculated using production logic.
99
+ - Every bug fix ships with a regression test.
100
+ - Every error has two audiences: the system (full diagnostics) and the consumer (only actionable info). Never the same message.
101
+ - Errors change shape at every module boundary. No error propagates without translation.
102
+ - Errors never reveal system internals to consumers. No stack traces, file paths, or queries in responses.
103
+ - Graceful degradation: live data → cached → static fallback → feature unavailable.
104
+ - Every input is hostile until validated.
105
+ - Default deny. Any permission not explicitly granted is denied.
106
+ - Secrets never logged, never in error messages, never in responses, never committed.
107
+ - Dependencies flow downward only. Never import from a layer above.
108
+ - Each external service has exactly one integration module that owns its boundary.
109
+ - Data crosses boundaries as plain values. Never pass ORM instances or SDK types between layers.
110
+ - ASCII diagrams for data flow, state machines, and architecture. Use box-drawing characters (─│┌┐└┘├┤┬┴┼) and arrows (→←↑↓).
111
+
112
+ ---
113
+
114
+ ## Shell Execution
115
+
116
+ Shell commands use Unix syntax (Git Bash). Never use CMD (`dir`, `type`, `del`) or backslash paths in Bash tool calls. On Windows, use forward slashes, `ls`, `grep`, `rm`, `cat`.
117
+
118
+ ---
119
+
120
+ ## AskUserQuestion
121
+
122
+ **Contract:**
123
+ 1. **Re-ground:** Project name, branch, current task. (1-2 sentences.)
124
+ 2. **Simplify:** Plain English a smart 16-year-old could follow.
125
+ 3. **Recommend:** Name the recommended option and why.
126
+ 4. **Options:** Ordered by completeness descending.
127
+ 5. **One decision per question.**
128
+
129
+ **When to ask (mandatory):**
130
+ 1. Design/UX choice not resolved in artifacts
131
+ 2. Trade-off with more than one viable option
132
+ 3. Before writing to files outside .warp/
133
+ 4. Deviating from architecture or design spec
134
+ 5. Skipping or deferring an acceptance criterion
135
+ 6. Before any destructive or irreversible action
136
+ 7. Ambiguous or underspecified requirement
137
+ 8. Choosing between competing library/tool options
138
+
139
+ **Completeness scores in labels (mandatory):**
140
+ Format: `"Option name — X/10 🟢"` (or 🟡 or 🔴). In the label, not the description.
141
+ Rate: 🟢 9-10 complete, 🟡 6-8 adequate, 🔴 1-5 shortcuts.
142
+
143
+ **Formatting:**
144
+ - *Italics* for emphasis, not **bold** (bold for headers only).
145
+ - After each answer: `✔ Decision {N} recorded [quicksave updated]`
146
+ - Previews under 8 lines. Full mockups go in conversation text before the question.
147
+
148
+ ---
149
+
150
+ ## Scale Detection
151
+
152
+ - **Feature:** One capability/screen/endpoint. Lean phases, fewer questions.
153
+ - **Module:** A package or subsystem. Full depth, multiple concerns.
154
+ - **System:** Whole product or greenfield. Maximum depth, every edge case.
155
+
156
+ Detection: Single behavior change → feature. 3+ files → module. Cross-package → system.
157
+
158
+ ---
159
+
160
+ ## Artifact I/O
161
+
162
+ Header: `<!-- Pipeline: {skill-name} | {date} | Scale: {scale} | Inputs: {prerequisites} -->`
163
+
164
+ Validation: all schema sections present, no empty sections, key decisions explicit.
165
+ Preview: show first 8-10 lines + total line count before writing.
166
+ HTML preview: use `_warp_html.sh` if available. Open in browser at hard gates only.
167
+
168
+ ---
169
+
170
+ ## Completion Banner
171
+
172
+ ```
173
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
174
+ WARP │ {skill-name} │ {STATUS}
175
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
176
+ Wrote: {artifact path(s)}
177
+ Decisions: {N} recorded
178
+ Next: /{next-skill}
179
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
180
+ ```
181
+
182
+ Status values: **DONE**, **DONE_WITH_CONCERNS** (list concerns), **BLOCKED** (state blocker + what was tried + next steps), **NEEDS_CONTEXT** (state exactly what's needed).
183
+
184
+ <!-- ═══════════════════════════════════════════════════════════ -->
185
+ <!-- Skill-Specific Content. -->
186
+ <!-- ═══════════════════════════════════════════════════════════ -->
187
+
188
+
189
+ # Upgrade
190
+
191
+ Meta skill. Upgrades WarpOS from inside Claude Code — no terminal switching.
192
+
193
+ **The upgrade is ONE COMMAND.** Do not decompose into manual steps. The upgrade
194
+ script handles pull, build, install, auto-patch, changelog, and migration detection.
195
+
196
+ ---
197
+
198
+ ## STEP 1: Locate the WarpOS Repo
199
+
200
+ ```bash
201
+ WARP_REPO=""
202
+ for candidate in \
203
+ "$(cat "$HOME/.warp/repo-path" 2>/dev/null)" \
204
+ "$(readlink -f ~/.claude/skills/warp-setup 2>/dev/null)/../../../.." \
205
+ "$(dirname "$(readlink -f ~/.warp/hooks/identity-briefing.sh 2>/dev/null)")/../.." \
206
+ "$HOME/Projects/warp" \
207
+ "$HOME/warp" \
208
+ ; do
209
+ [ -z "$candidate" ] && continue
210
+ if [ -d "$candidate/.git" ] && [ -f "$candidate/VERSION" ] && [ -f "$candidate/build.sh" ]; then
211
+ WARP_REPO="$(cd "$candidate" && pwd)"
212
+ break
213
+ fi
214
+ done
215
+ echo "Warp repo: ${WARP_REPO:-NOT FOUND}"
216
+ ```
217
+
218
+ If not found, ask the user for the path.
219
+
220
+ ---
221
+
222
+ ## STEP 2: Run the Upgrade Script
223
+
224
+ **Run `bin/warp-upgrade.sh` as a single command. Do NOT decompose into manual steps.**
225
+
226
+ The script handles everything: fetch, version check, pull, build, install, auto-patch
227
+ project hooks, changelog display, migration detection, and restart recommendation.
228
+
229
+ ```bash
230
+ cd "$WARP_REPO" && bash bin/warp-upgrade.sh
231
+ ```
232
+
233
+ **IMPORTANT — Timeout:** On Windows (Git Bash), build + install can take 2-4 minutes.
234
+ Use a Bash timeout of at least 300000ms (5 minutes). If the default timeout is shorter,
235
+ set it explicitly. A timed-out upgrade leaves a half-installed state.
236
+
237
+ **If the script succeeds:** Read its output, summarize what changed, and remind the user
238
+ to restart Claude Code.
239
+
240
+ **If the script fails:** Read the error output. Common failures:
241
+ - Network error on git pull → "Check your connection and try again"
242
+ - Build error → "build.sh failed — check the error above"
243
+ - Install error (Windows file locks) → "Close other Claude Code instances and retry"
244
+ - Git conflicts → "Your local Warp repo has changes. Run `git stash` first."
245
+
246
+ Only if the script itself is missing or broken should you fall back to manual steps.
247
+
248
+ ---
249
+
250
+ ## STEP 3: Verify Auto-Patch (if script completed)
251
+
252
+ The upgrade script auto-patches `.claude/settings.local.json` in the current project
253
+ directory with the latest hook config from `~/.warp/project-hooks.json`.
254
+
255
+ Verify the patch applied:
256
+
257
+ ```bash
258
+ # Check that settings.local.json references new hook names
259
+ grep -c 'identity-foundation\|identity-briefing\|consistency-check' .claude/settings.local.json 2>/dev/null && echo "hooks patched" || echo "hooks NOT patched"
260
+ ```
261
+
262
+ If NOT patched (script ran from a different directory, or no settings.local.json):
263
+ - If `settings.local.json` exists but has old hook names → patch it manually using
264
+ `~/.warp/project-hooks.json` as the source. Preserve `permissions` and `enabledMcpjsonServers`.
265
+ - If no `settings.local.json` → tell user to run `/warp-setup` (first-time setup, not upgrade).
266
+
267
+ ---
268
+
269
+ ## FALLBACK: Manual Steps (only if warp-upgrade.sh is missing/broken)
270
+
271
+ If `bin/warp-upgrade.sh` does not exist or crashes on invocation, fall back to manual:
272
+
273
+ ```bash
274
+ cd "$WARP_REPO"
275
+ git fetch origin --quiet
276
+ git pull --rebase origin master
277
+ bash build.sh
278
+ bash install.sh
279
+ cat CHANGELOG.md | head -60
280
+ ```
281
+
282
+ Then manually patch project hooks:
283
+ ```bash
284
+ # Read template and patch settings.local.json
285
+ python3 -c "
286
+ import json
287
+ t = json.load(open('$HOME/.warp/project-hooks.json'))
288
+ s = json.load(open('.claude/settings.local.json'))
289
+ s['agent'] = t['agent']
290
+ s['hooks'] = t['hooks']
291
+ json.dump(s, open('.claude/settings.local.json','w'), indent=2)
292
+ print('Patched.')
293
+ "
294
+ ```
295
+
296
+ **This fallback exists for emergencies only.** The script is the primary path.
297
+
298
+ ---
299
+
300
+ ## OUTPUT FORMAT
301
+
302
+ ```
303
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
304
+ WARP │ UPGRADE
305
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
306
+ Repo: {path}
307
+ Upgrade: v{old} → v{new}
308
+
309
+ {changelog excerpt — Added/Changed/Removed}
310
+
311
+ Hooks: auto-patched ✓
312
+ Restart: required (new hooks + skills)
313
+ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
314
+ ```
315
+
316
+ **Always remind:** "Restart Claude Code to pick up the new skills and hooks."
317
+
318
+ ---
319
+
320
+ ## MUST / MUST NOT
321
+
322
+ **MUST:**
323
+ - Run `bin/warp-upgrade.sh` as a single command (not decomposed steps)
324
+ - Use a 5-minute timeout for the upgrade script on Windows
325
+ - Verify hooks auto-patched after script completes
326
+ - Show changelog summary
327
+ - Remind user to restart Claude Code
328
+
329
+ **MUST NOT:**
330
+ - Decompose the upgrade into individual git/build/install commands when the script works
331
+ - Use default Bash timeout (2 minutes is too short on Windows)
332
+ - Proceed to install if build fails
333
+ - Require the user to re-run `/warp-setup` after a normal upgrade
334
+ - Force-push, reset, or destructively modify the Warp repo without consent
package/bin/cli.js ADDED
@@ -0,0 +1,44 @@
1
+ #!/usr/bin/env node
2
+ // cli.js — Warp CLI entry point
3
+ // Usage: npx warp-os install | upgrade | version | uninstall
4
+
5
+ const { execSync } = require('child_process');
6
+ const path = require('path');
7
+
8
+ const command = process.argv[2];
9
+
10
+ if (!command || command === '--help' || command === '-h') {
11
+ console.log(`
12
+ warp-os — A development operating system for Claude Code
13
+
14
+ Usage:
15
+ npx warp-os install Install Warp (skills, agents, hooks, claude-mem)
16
+ npx warp-os upgrade Upgrade to the latest version
17
+ npx warp-os uninstall Remove Warp from this machine
18
+ npx warp-os version Show installed version
19
+ npx warp-os doctor Check installation health
20
+ `);
21
+ process.exit(0);
22
+ }
23
+
24
+ if (command === 'version' || command === '--version' || command === '-v') {
25
+ const version = require('../package.json').version;
26
+ console.log(`warp-os v${version}`);
27
+ process.exit(0);
28
+ }
29
+
30
+ if (command === 'install' || command === 'upgrade') {
31
+ require('./install.js');
32
+ } else if (command === 'uninstall') {
33
+ console.log('Uninstall not yet implemented. Remove manually:');
34
+ console.log(' rm -rf ~/.claude/skills/warp-*');
35
+ console.log(' rm -rf ~/.claude/agents/warp-*.md');
36
+ console.log(' rm -rf ~/.warp/');
37
+ process.exit(0);
38
+ } else if (command === 'doctor') {
39
+ require('./install.js');
40
+ } else {
41
+ console.error(`Unknown command: ${command}`);
42
+ console.error('Run "npx warp-os --help" for usage.');
43
+ process.exit(1);
44
+ }
@@ -0,0 +1,291 @@
1
+ #!/usr/bin/env bash
2
+ # _warp_html.sh — Shared utility for markdown-to-HTML conversion
3
+ # Source this file in skills/hooks that need HTML preview generation.
4
+ # Provides:
5
+ # _warp_md_to_html "input.md" "artifact_name" "output.html"
6
+ # _warp_open_preview "file.html"
7
+ #
8
+ # Uses the warp streak CSS palette. Sanitizes HTML tags to prevent XSS (F-1).
9
+ # No external dependencies — pure bash + sed.
10
+
11
+ # Convert markdown to HTML with warp streak CSS
12
+ # Usage: _warp_md_to_html "input.md" "Artifact Name" "output.html"
13
+ _warp_md_to_html() {
14
+ local input="$1"
15
+ local artifact_name="$2"
16
+ local output="$3"
17
+
18
+ if [ ! -f "$input" ]; then
19
+ echo "[WARP WARNING] _warp_md_to_html: input file not found: $input" >&2
20
+ return 1
21
+ fi
22
+
23
+ # Ensure output directory exists
24
+ mkdir -p "$(dirname "$output")" 2>/dev/null || true
25
+
26
+ # Convert markdown to HTML with state tracking for code blocks and comments
27
+ local body=""
28
+ local in_code=false
29
+ local in_comment=false
30
+
31
+ while IFS= read -r line || [ -n "$line" ]; do
32
+ # --- HTML comment handling: skip entirely ---
33
+ if [[ "$line" == *"<!--"* ]]; then
34
+ if [[ "$line" == *"-->"* ]]; then
35
+ continue # single-line comment — skip
36
+ fi
37
+ in_comment=true
38
+ continue
39
+ fi
40
+ if $in_comment; then
41
+ if [[ "$line" == *"-->"* ]]; then
42
+ in_comment=false
43
+ fi
44
+ continue
45
+ fi
46
+
47
+ # --- Fenced code blocks ---
48
+ if [[ "$line" =~ ^\`\`\` ]]; then
49
+ if $in_code; then
50
+ body+="</code></pre>"$'\n'
51
+ in_code=false
52
+ else
53
+ body+="<pre><code>"
54
+ in_code=true
55
+ fi
56
+ continue
57
+ fi
58
+
59
+ if $in_code; then
60
+ # Inside code block: escape HTML but preserve whitespace exactly
61
+ line="${line//&/&amp;}"
62
+ line="${line//</&lt;}"
63
+ line="${line//>/&gt;}"
64
+ body+="$line"$'\n'
65
+ continue
66
+ fi
67
+
68
+ # --- Normal markdown conversion (outside code blocks) ---
69
+ # Escape HTML entities
70
+ line="${line//&/&amp;}"
71
+ line="${line//</&lt;}"
72
+ line="${line//>/&gt;}"
73
+
74
+ # Headings
75
+ if [[ "$line" =~ ^###\ ]]; then
76
+ line="<h3>${line#\#\#\# }</h3>"
77
+ elif [[ "$line" =~ ^##\ ]]; then
78
+ line="<h2>${line#\#\# }</h2>"
79
+ elif [[ "$line" =~ ^#\ ]]; then
80
+ line="<h1>${line#\# }</h1>"
81
+ # Horizontal rule
82
+ elif [[ "$line" == "---" ]]; then
83
+ line="<hr>"
84
+ # List items
85
+ elif [[ "$line" =~ ^-\ ]]; then
86
+ line="<li>${line#- }</li>"
87
+ elif [[ "$line" =~ ^[0-9]+\.\ ]]; then
88
+ line="<li>${line#*. }</li>"
89
+ # Blank line
90
+ elif [[ -z "$line" ]]; then
91
+ line="<br>"
92
+ # Table rows (pipe-delimited)
93
+ elif [[ "$line" == *"|"* ]] && [[ "$line" =~ ^\|.*\|$ ]]; then
94
+ # Skip separator rows (|---|---|)
95
+ if [[ "$line" =~ ^[[:space:]]*\|[-[:space:]|]+\|[[:space:]]*$ ]]; then
96
+ continue
97
+ fi
98
+ local cells=""
99
+ local tag="td"
100
+ # Simple heuristic: if all cells are bold, use th
101
+ if [[ "$line" == *"**"* ]] && [[ $(echo "$line" | grep -o '\*\*' | wc -l) -ge 4 ]]; then
102
+ tag="th"
103
+ fi
104
+ IFS='|' read -ra PARTS <<< "$line"
105
+ for part in "${PARTS[@]}"; do
106
+ part=$(echo "$part" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')
107
+ [ -z "$part" ] && continue
108
+ cells+="<$tag>$part</$tag>"
109
+ done
110
+ line="<tr>$cells</tr>"
111
+ fi
112
+
113
+ # Inline formatting (outside headings processing)
114
+ line=$(echo "$line" | sed 's/\*\*\([^*]*\)\*\*/<strong>\1<\/strong>/g')
115
+ line=$(echo "$line" | sed 's/\*\([^*]*\)\*/<em>\1<\/em>/g')
116
+ line=$(echo "$line" | sed 's/`\([^`]*\)`/<code>\1<\/code>/g')
117
+
118
+ body+="$line"$'\n'
119
+ done < "$input"
120
+
121
+ # Close unclosed code block
122
+ if $in_code; then
123
+ body+="</code></pre>"$'\n'
124
+ fi
125
+
126
+ # Wrap consecutive <tr> rows in <table> tags
127
+ body=$(echo "$body" | awk '
128
+ /<tr>/ && !intable { print "<table>"; intable=1 }
129
+ !/<tr>/ && intable { print "</table>"; intable=0 }
130
+ { print }
131
+ END { if (intable) print "</table>" }
132
+ ')
133
+
134
+ # Write the full HTML file
135
+ cat > "$output" << HTMLEOF
136
+ <!DOCTYPE html>
137
+ <html lang="en">
138
+ <head>
139
+ <meta charset="utf-8">
140
+ <meta name="viewport" content="width=device-width, initial-scale=1">
141
+ <title>Warp │ ${artifact_name}</title>
142
+ <style>
143
+ :root {
144
+ --space-deep: #0d0d14;
145
+ --space-surface: #1a1528;
146
+ --uv-border: #2d2444;
147
+ --uv-glow: #6b5ce7;
148
+ --streak-primary: #e8e6f0;
149
+ --streak-bright: #f5f4f8;
150
+ --streak-dim: #a9a5b8;
151
+ --flare-cyan: #58e8ff;
152
+ --flare-magenta: #ff58e4;
153
+ --flare-amber: #ffc858;
154
+ --flare-red: #ff5878;
155
+ --flare-green: #58ff9e;
156
+ }
157
+ * { margin: 0; padding: 0; box-sizing: border-box; }
158
+ body {
159
+ font: 16px/1.8 -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
160
+ background: var(--space-deep); color: var(--streak-primary);
161
+ max-width: 860px; margin: 0 auto; padding: 0 1.5rem 4rem;
162
+ /* Subtle radial glow from top center */
163
+ background-image: radial-gradient(ellipse 80% 50% at 50% -5%, rgba(107, 92, 231, 0.12) 0%, transparent 70%);
164
+ background-attachment: fixed;
165
+ }
166
+ /* Hero header — full-width gradient bar with glow */
167
+ .warp-header {
168
+ font: 700 0.8rem/1 'JetBrains Mono', 'Fira Code', monospace;
169
+ letter-spacing: 0.2em; text-transform: uppercase;
170
+ padding: 1.5rem 0;
171
+ margin: 0 -1.5rem 3rem;
172
+ padding-left: 1.5rem;
173
+ border-bottom: 2px solid transparent;
174
+ border-image: linear-gradient(90deg, var(--flare-cyan), var(--flare-magenta), var(--flare-amber), var(--flare-green), var(--flare-cyan)) 1;
175
+ background: linear-gradient(180deg, rgba(107, 92, 231, 0.08) 0%, transparent 100%);
176
+ color: var(--streak-dim);
177
+ }
178
+ .warp-header span {
179
+ background: linear-gradient(90deg, var(--flare-cyan), var(--flare-magenta), var(--flare-amber), var(--flare-green));
180
+ -webkit-background-clip: text; -webkit-text-fill-color: transparent;
181
+ background-clip: text;
182
+ }
183
+ /* H1 — large with rainbow underline accent */
184
+ h1 {
185
+ color: var(--streak-bright); font-size: 1.6rem; font-weight: 700;
186
+ margin: 2.5rem 0 1.2rem; padding-bottom: 0.6rem;
187
+ border-bottom: 2px solid transparent;
188
+ border-image: linear-gradient(90deg, var(--flare-cyan), var(--flare-magenta), transparent 80%) 1;
189
+ }
190
+ /* H2 — section cards */
191
+ h2 {
192
+ color: var(--streak-bright); font-size: 1.15rem; font-weight: 600;
193
+ margin: 2.5rem -1rem 1rem; padding: 0.7rem 1rem;
194
+ background: linear-gradient(135deg, rgba(107, 92, 231, 0.1) 0%, rgba(26, 21, 40, 0.6) 100%);
195
+ border-left: 3px solid var(--uv-glow);
196
+ border-radius: 0 8px 8px 0;
197
+ }
198
+ h3 {
199
+ color: var(--streak-bright); font-size: 1rem; font-weight: 600;
200
+ margin: 1.8rem 0 0.5rem; padding-left: 0.8rem;
201
+ border-left: 2px solid var(--uv-glow);
202
+ }
203
+ p { margin: 0.7rem 0; }
204
+ a { color: var(--flare-cyan); text-decoration: none; border-bottom: 1px solid rgba(88, 232, 255, 0.3);
205
+ transition: border-color 0.2s; }
206
+ a:hover { border-bottom-color: var(--flare-cyan); }
207
+ strong { color: var(--streak-bright); }
208
+ em { color: var(--streak-dim); font-style: italic; }
209
+ /* Inline code — pill with glow */
210
+ code {
211
+ font-family: 'JetBrains Mono', 'Fira Code', monospace;
212
+ background: rgba(107, 92, 231, 0.1); padding: 2px 8px; border-radius: 4px;
213
+ font-size: 0.86em; border: 1px solid rgba(107, 92, 231, 0.25);
214
+ color: var(--streak-bright);
215
+ }
216
+ /* Code blocks — elevated card with UV border glow */
217
+ pre {
218
+ background: linear-gradient(180deg, var(--space-surface) 0%, rgba(13, 13, 20, 0.95) 100%);
219
+ padding: 1.4rem 1.6rem; border-radius: 10px;
220
+ border: 1px solid var(--uv-border); overflow-x: auto; margin: 1.4rem 0;
221
+ box-shadow: 0 8px 32px rgba(0, 0, 0, 0.4), 0 0 1px rgba(107, 92, 231, 0.3);
222
+ position: relative;
223
+ }
224
+ pre::before {
225
+ content: ''; position: absolute; top: -1px; left: 20%; right: 20%; height: 1px;
226
+ background: linear-gradient(90deg, transparent, var(--uv-glow), transparent);
227
+ }
228
+ pre code { background: none; padding: 0; font-size: 0.84em; line-height: 1.6;
229
+ border: none; color: var(--streak-primary); }
230
+ /* Tables — frosted glass effect */
231
+ table {
232
+ border-collapse: separate; border-spacing: 0; width: 100%; margin: 1.4rem 0;
233
+ font-size: 0.92em; border-radius: 8px; overflow: hidden;
234
+ border: 1px solid var(--uv-border);
235
+ box-shadow: 0 4px 16px rgba(0, 0, 0, 0.2);
236
+ }
237
+ th, td { border-bottom: 1px solid var(--uv-border); padding: 0.6rem 0.85rem; text-align: left; }
238
+ td { border-right: 1px solid rgba(45, 36, 68, 0.5); }
239
+ td:last-child { border-right: none; }
240
+ th {
241
+ background: rgba(107, 92, 231, 0.12);
242
+ color: var(--streak-bright); font-weight: 600; font-size: 0.8em;
243
+ text-transform: uppercase; letter-spacing: 0.06em;
244
+ border-bottom: 2px solid var(--uv-border);
245
+ }
246
+ tr:last-child td { border-bottom: none; }
247
+ tr:hover td { background: rgba(107, 92, 231, 0.06); transition: background 0.15s; }
248
+ blockquote {
249
+ border-left: 3px solid var(--uv-glow); padding: 0.8rem 1.2rem;
250
+ color: var(--streak-dim); margin: 1.2rem 0;
251
+ background: rgba(107, 92, 231, 0.04); border-radius: 0 8px 8px 0;
252
+ }
253
+ li { margin: 0.35rem 0; margin-left: 1.5rem; }
254
+ li::marker { color: var(--uv-glow); }
255
+ /* HR — rainbow streak */
256
+ hr {
257
+ border: none; height: 1px; margin: 3rem 0;
258
+ background: linear-gradient(90deg, transparent 5%, var(--flare-cyan) 20%, var(--flare-magenta) 50%, var(--flare-amber) 80%, transparent 95%);
259
+ opacity: 0.4;
260
+ }
261
+ /* Scrollbar */
262
+ ::-webkit-scrollbar { width: 6px; height: 6px; }
263
+ ::-webkit-scrollbar-track { background: var(--space-deep); }
264
+ ::-webkit-scrollbar-thumb { background: var(--uv-border); border-radius: 3px; }
265
+ ::-webkit-scrollbar-thumb:hover { background: var(--uv-glow); }
266
+ </style>
267
+ </head>
268
+ <body>
269
+ <div class="warp-header"><span>Warp</span> │ ${artifact_name}</div>
270
+ ${body}
271
+ </body>
272
+ </html>
273
+ HTMLEOF
274
+
275
+ # Set restrictive permissions (security.md recommendation)
276
+ chmod 600 "$output" 2>/dev/null || true
277
+ }
278
+
279
+ # Open a file in the user's default browser (cross-platform)
280
+ # Usage: _warp_open_preview "file.html"
281
+ _warp_open_preview() {
282
+ local file="$1"
283
+ if [ ! -f "$file" ]; then
284
+ return 0
285
+ fi
286
+ case "$(uname -s)" in
287
+ MINGW*|MSYS*|CYGWIN*) start "" "$file" 2>/dev/null || true ;;
288
+ Darwin) open "$file" 2>/dev/null || true ;;
289
+ *) xdg-open "$file" 2>/dev/null || true ;;
290
+ esac
291
+ }