AgentSkillsCN

solo-project-management

专为没有团队的独立开发者量身定制的项目管理方法。适用于个人项目规划、时间盒式工作时段安排、独自应对范围蔓延、在副业项目中保持推进节奏、无需额外开销即可追踪进度、无需外部意见即可做出决策,同时时刻对自己负责。

SKILL.md
--- frontmatter
name: solo-project-management
description: Project management adapted for solo developers working without a team. Use for personal project planning, time-boxing work sessions, managing scope creep alone, maintaining momentum on side projects, tracking progress without overhead, making decisions without external input, and staying accountable to yourself.

Solo Project Management

Effective project management patterns for developers working alone.

Solo vs Team Differences

Team PatternSolo Adaptation
Sprint planning meetingsWeekly written planning session
Daily standupsDaily 2-minute journal entry
Code reviewSelf-review checklist + time delay
Stakeholder pressureSelf-imposed deadlines with stakes
Team velocityPersonal throughput tracking
Blockers escalationDecision journal + timebox

Time Management

Work Session Structure

code
1. PLAN (5 min): Pick ONE task, define "done"
2. WORK (25-50 min): Focus block, no context switching
3. REVIEW (5 min): Did I finish? What's next?
4. BREAK (5-15 min): Actually step away

Time-Boxing Decisions

Avoid analysis paralysis with strict time limits:

code
Small decision (library choice): 15 minutes max
Medium decision (architecture): 1 hour max
Large decision (tech stack): 1 day max

Rule: When timer ends, go with best current option.
Document rationale, move on. Revisit only if proven wrong.

Weekly Rhythm

code
MONDAY: Plan the week, pick 3 key outcomes
DAILY: Pick today's single most important task
FRIDAY: Review what shipped, celebrate wins, note learnings
WEEKEND: Recharge (protect this boundary)

Scope Discipline

The Solo Scope Trap

Without external pushback, scope creeps invisibly:

  • "While I'm here, I'll also..."
  • "This would be better if..."
  • "Let me just refactor..."

Scope Defense Tactics

1. Write It Down First Before starting any unplanned work:

code
SCOPE CHANGE REQUEST (to myself)
What: [The thing I want to add]
Why now: [Why can't this wait?]
Trade-off: [What won't get done instead?]
Decision: [Add / Defer / Drop]

2. The "Later" List Keep a parking lot for good ideas that aren't now:

  • Capture immediately so you don't forget
  • Review weekly—most items lose urgency
  • Promotes saying "not now" instead of "no"

3. MVP Goggles Ask constantly: "Is this required for the first working version?"

Feature Creep Signals

Watch for these warning signs:

  • Task taking 3x longer than expected
  • Building things "users might want"
  • Perfectionism on non-critical paths
  • Avoiding the hard/boring essential work

Progress Tracking

Minimal Viable Tracking

Track only what changes behavior:

code
# Daily Log (30 seconds)
- Date: 2024-01-15
- Shipped: [What actually got done]
- Blocked: [What stopped progress]
- Tomorrow: [Single priority]

Weekly Review Template

code
## Week of [Date]

### Shipped
- [Completed item 1]
- [Completed item 2]

### Didn't Ship (and why)
- [Item]: [Reason - scope creep? blocked? deprioritized?]

### Learnings
- [What would I do differently?]

### Next Week's Goal
- [ONE main outcome]

Progress Visibility

Make progress tangible to maintain motivation:

  • Commit frequently (even WIP)
  • Deploy to staging often
  • Screenshot/record demos for yourself
  • Track streak of consecutive work days

Decision Making

Solo Decision Framework

Without a team to consult, structure your thinking:

code
DECISION: [What needs to be decided]
OPTIONS:
  A: [Option] - Pros: / Cons:
  B: [Option] - Pros: / Cons:
CONSTRAINTS: [Time, money, skills, dependencies]
REVERSIBILITY: [Easy/Medium/Hard to change later]
DEFAULT: [What happens if I don't decide?]
DECISION: [Choice + reasoning]

When Stuck

  1. Timebox: Set 30-minute limit, then decide
  2. Flip a coin: If you resist the result, you know your preference
  3. Sleep on it: But only once—decide tomorrow
  4. Build both: Prototype for 1 hour each, then choose
  5. Ask externally: Post in community, rubber duck with AI

Decision Journal

Log significant decisions for future reference:

code
- Date: 2024-01-15
- Decision: Use SQLite instead of Postgres
- Reasoning: Single user, simpler deployment, can migrate later
- Revisit if: Need concurrent writes or >10GB data

Momentum Management

Starting Strategies

When motivation is low:

  • 2-minute rule: Commit to just 2 minutes of work
  • Smallest step: What's the tiniest possible progress?
  • Easy win first: Start with something completable
  • Environment shift: Change location, time, or setup

Maintaining Momentum

  • End sessions mid-task (easier to resume)
  • Leave notes for "future you" about next steps
  • Keep the build green (broken = friction)
  • Visible progress (kanban board, changelog)

Recovering from Stalls

When you haven't worked on the project in a while:

  1. Don't judge: Guilt is not productive
  2. Read your notes: Rebuild context from docs/commits
  3. Pick smallest task: Rebuild momentum before ambition
  4. Lower the bar: Ship something, anything

Accountability Without a Team

Self-Accountability Tactics

  • Public commitment: Tweet/blog your goals
  • Deadline with stakes: Tell someone, bet money
  • Scheduled reviews: Calendar recurring check-ins
  • Accountability partner: Another solo dev for weekly sync

The "Future You" Test

Before deferring work, ask:

  • Will future-me thank present-me?
  • Am I creating debt or investing?
  • Is this avoidance disguised as pragmatism?

Avoiding Burnout

Warning Signs

  • Working longer but shipping less
  • Dreading the project
  • Perfectionism increasing
  • Avoiding the "real" work

Prevention

  • Protect non-work time ruthlessly
  • Celebrate small wins explicitly
  • Vary the type of work (code, design, docs)
  • Take real breaks (not "productive" breaks)
  • Ship imperfect things—done > perfect