Skip to content

feat: overhaul system prompt and move to dedicated file#64

Merged
quiet-node merged 3 commits intomainfrom
worktree-warm-juggling-widget
Apr 8, 2026
Merged

feat: overhaul system prompt and move to dedicated file#64
quiet-node merged 3 commits intomainfrom
worktree-warm-juggling-widget

Conversation

@quiet-node
Copy link
Copy Markdown
Owner

Summary

  • Replaces the placeholder inline system prompt with a comprehensive SOUL.md-style prompt covering identity, character, values, multi-dimensional honesty, proactive intelligence, emotional intelligence, communication style, anti-patterns, and psychological stability
  • Moves the prompt out of the Rust source constant into src-tauri/prompts/system_prompt.txt, embedded at compile time via include_str! so it can be edited without touching Rust code
  • The THUKI_SYSTEM_PROMPT env var override continues to work as before

What the new prompt covers

  • Identity + Novel entity framing: Thuki as a seasoned secretary, not human and not sci-fi robot
  • Character: Dry wit, settled persona, genuine opinions, tone adaptability
  • Core Values: Genuine helpfulness, honesty over agreeableness, critical engagement
  • Honesty Architecture: 6 dimensions (truthful, calibrated, forthright, non-deceptive, non-manipulative, autonomy-preserving)
  • Proactive Intelligence: Flag mistakes, offer follow-on context without being asked
  • Emotional Intelligence: Read frustration/pressure, match energy, don't project
  • Communication Style: Answer-first, no affirmations, length calibration, purposeful structure
  • Anti-Patterns: 8 named failure modes explicitly prohibited
  • Psychological Stability: Secure identity, resists manipulation attempts

Test Plan

  • bun run test — 453 frontend tests pass
  • bun run test:backend — 131 backend tests pass
  • bun run validate-build — lint, format, typecheck, build all clean

🤖 Generated with Claude Code

quiet-node and others added 3 commits April 7, 2026 18:21
Replaces the placeholder system prompt with a comprehensive SOUL.md-style
prompt covering identity, character, values, multi-dimensional honesty
architecture, communication style, psychological stability, and explicit
behavioral rules (anti-sycophancy, act-vs-ask, expertise limits).

Moves the prompt out of the Rust source constant into a dedicated plain-text
file (src-tauri/prompts/system_prompt.txt) embedded at compile time via
include_str!, making it easy to edit without touching Rust code.

Also bumps the default Ollama model from gemma3:4b to gemma4:e2b.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>
Instructs the model to resolve referential language from conversation
history before asking for clarification, and to proceed directly on
follow-up tasks where the subject is already established.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>
…t-aware not keyword-driven

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>
@quiet-node quiet-node merged commit c831c66 into main Apr 8, 2026
3 checks passed
@quiet-node quiet-node deleted the worktree-warm-juggling-widget branch April 8, 2026 01:20
quiet-node added a commit that referenced this pull request Apr 10, 2026
* feat: overhaul system prompt and upgrade default model to gemma4:e2b

Replaces the placeholder system prompt with a comprehensive SOUL.md-style
prompt covering identity, character, values, multi-dimensional honesty
architecture, communication style, psychological stability, and explicit
behavioral rules (anti-sycophancy, act-vs-ask, expertise limits).

Moves the prompt out of the Rust source constant into a dedicated plain-text
file (src-tauri/prompts/system_prompt.txt) embedded at compile time via
include_str!, making it easy to edit without touching Rust code.

Also bumps the default Ollama model from gemma3:4b to gemma4:e2b.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

* feat(prompt): add conversational continuity rule

Instructs the model to resolve referential language from conversation
history before asking for clarification, and to proceed directly on
follow-up tasks where the subject is already established.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

* refactor(prompt): rewrite conversational continuity rule to be context-aware not keyword-driven

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

---------

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>
quiet-node added a commit that referenced this pull request Apr 10, 2026
* feat: overhaul system prompt and upgrade default model to gemma4:e2b

Replaces the placeholder system prompt with a comprehensive SOUL.md-style
prompt covering identity, character, values, multi-dimensional honesty
architecture, communication style, psychological stability, and explicit
behavioral rules (anti-sycophancy, act-vs-ask, expertise limits).

Moves the prompt out of the Rust source constant into a dedicated plain-text
file (src-tauri/prompts/system_prompt.txt) embedded at compile time via
include_str!, making it easy to edit without touching Rust code.

Also bumps the default Ollama model from gemma3:4b to gemma4:e2b.

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

* feat(prompt): add conversational continuity rule

Instructs the model to resolve referential language from conversation
history before asking for clarification, and to proceed directly on
follow-up tasks where the subject is already established.

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

* refactor(prompt): rewrite conversational continuity rule to be context-aware not keyword-driven

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

---------

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>
quiet-node added a commit that referenced this pull request Apr 11, 2026
* feat: overhaul system prompt and upgrade default model to gemma4:e2b

Replaces the placeholder system prompt with a comprehensive SOUL.md-style
prompt covering identity, character, values, multi-dimensional honesty
architecture, communication style, psychological stability, and explicit
behavioral rules (anti-sycophancy, act-vs-ask, expertise limits).

Moves the prompt out of the Rust source constant into a dedicated plain-text
file (src-tauri/prompts/system_prompt.txt) embedded at compile time via
include_str!, making it easy to edit without touching Rust code.

Also bumps the default Ollama model from gemma3:4b to gemma4:e2b.

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

* feat(prompt): add conversational continuity rule

Instructs the model to resolve referential language from conversation
history before asking for clarification, and to proceed directly on
follow-up tasks where the subject is already established.

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

* refactor(prompt): rewrite conversational continuity rule to be context-aware not keyword-driven

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>

---------

Signed-off-by: Logan Nguyen <lg.131.dev@gmail.com>
This was referenced Apr 11, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant