deepflow 0.1.18 → 0.1.19

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/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "deepflow",
3
- "version": "0.1.18",
3
+ "version": "0.1.19",
4
4
  "description": "Stay in flow state - lightweight spec-driven task orchestration for Claude Code",
5
5
  "keywords": [
6
6
  "claude",
@@ -86,20 +86,47 @@ Include patterns in task descriptions for agents to follow.
86
86
  2. Impact — core features before enhancements
87
87
  3. Risk — unknowns early
88
88
 
89
- ### 5. OUTPUT PLAN.md
89
+ ### 5. VALIDATE HYPOTHESES
90
90
 
91
- Append tasks grouped by `### doing-{spec-name}`. Include spec gaps if any.
91
+ Before finalizing the plan, identify and test risky assumptions:
92
92
 
93
- ### 6. RENAME SPECS
93
+ **When to validate:**
94
+ - Unfamiliar APIs or libraries
95
+ - Architectural decisions with multiple approaches
96
+ - Integration with external systems
97
+ - Performance-critical paths
98
+
99
+ **How to validate:**
100
+ 1. Create minimal prototype (scratchpad, not committed)
101
+ 2. Test the specific assumption
102
+ 3. Document findings in task description
103
+ 4. Adjust approach if hypothesis fails
104
+
105
+ **Examples:**
106
+ - "Does SessionStart hook run once per session?" → Test with simple log
107
+ - "Can we use streaming for large files?" → Prototype with sample data
108
+ - "Will this regex handle edge cases?" → Test against real samples
109
+
110
+ **Skip validation when:**
111
+ - Using well-known patterns
112
+ - Simple CRUD operations
113
+ - Clear documentation exists
114
+
115
+ ### 6. OUTPUT PLAN.md
116
+
117
+ Append tasks grouped by `### doing-{spec-name}`. Include spec gaps and validation findings.
118
+
119
+ ### 7. RENAME SPECS
94
120
 
95
121
  `mv specs/feature.md specs/doing-feature.md`
96
122
 
97
- ### 7. REPORT
123
+ ### 8. REPORT
98
124
 
99
125
  `✓ Plan generated — {n} specs, {n} tasks. Run /df:execute`
100
126
 
101
127
  ## Rules
102
- - **Plan only** — Do NOT implement anything
128
+ - **Plan only** — Do NOT implement anything (except quick validation prototypes)
129
+ - **Validate before commit** — Test risky assumptions with minimal experiments
103
130
  - **Confirm before assume** — Search code before marking "missing"
104
131
  - **One task = one logical unit** — Atomic, committable
105
132
  - Prefer existing utilities over new code