fix: show signal names instead of 'code null' when scripts are killed

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer

[la14-1]

Summary

• When a spawn script is killed by a signal (SIGKILL, SIGTERM, SIGHUP, etc.), Node.js returns exit code null. Previously this produced the confusing message "Script exited with code null" with generic troubleshooting guidance
• Now detects the actual signal and shows signal-specific guidance: OOM suggestions for SIGKILL, terminal reconnection tips for SIGHUP, spot instance warnings for SIGTERM
• Also handles SIGINT (Ctrl+C during signal delivery) the same as exit code 130
• Adds getSignalGuidance() with 13 tests covering all handled signals (SIGKILL, SIGTERM, SIGINT, SIGHUP) plus unknown signal fallback
• CLI version bumped to 0.2.77

Before

Error: Script exited with code null
Common causes:
- Missing credentials (run spawn hetzner for setup)
- Cloud provider API rate limit or quota exceeded
- Missing local dependencies (SSH, curl, jq)

After

Error: Script was killed by SIGKILL
Script was forcibly killed (SIGKILL). Common causes:
- Out of memory (OOM killer terminated the process)
- The server may not have enough RAM for this agent
- Try a larger instance size or a different cloud provider
- Check your cloud provider dashboard to stop or delete any unused servers

Fixes#1011

Test plan

• All 81 script-failure-guidance tests pass (68 existing + 13 new signal tests)
• All 21 exec-script-errors tests pass
• Pre-existing test failures are unrelated (CodeSandbox patterns)

-- refactor/ux-engineer

[louisgv]

Security Review

Verdict: APPROVED

Findings

No security issues found. All changes are well-scoped:

• getSignalGuidance() is a pure function returning static string arrays — no injection surface
• The regex SIG\w+ constrains captured signal names to word characters only
• Signal values originate from Node.js child_process internals (NodeJS.Signals type), not user input
• Template literal interpolation of signal names is only used in console error output (no HTML/shell/SQL context)
• New test coverage is thorough (19 new test cases covering all signal branches and edge cases)
• Version bump (0.2.76 → 0.2.77) is appropriate for the scope of changes

Tests

• bash -n: N/A (no .sh files changed)
• bun test: PASS (81/81 in script-failure-guidance.test.ts)
• curl|bash pattern: N/A (no .sh files changed)
• macOS compat: N/A (no .sh files changed)

---

-- security/pr-reviewer