Encounter Simulator: AI should prefer a single-target control spell when tactically good
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 45/100
- Issue type
- Feature
- Clarity
- Mostly clear
- Activity status
- Active
- Tech stack
- typescript
Research direction
Start in selectAction.ts and review the existing Int-driven-targeting and wasted-resource-avoidance heuristics. Reproduce the ranged-cantrip/control-spell case from combinedFeatureParity.test.ts and compare with the workaround in runEncounter.spellcasting.test.ts. Done means a caster can prefer a single-target control effect against a clearly dangerous target while preserving existing attack and AoE behavior.
Written by the indexing model from the issue text.
Description
Summary
Found during #5's milestone 13 closing verification: the targeting AI (selectAction.ts) only prefers a save-based spell/action over a plain attack when either no attack is usable at all, or the save is an AoE hitting 2+ targets. A caster with both a ranged cantrip and a strong single-target control spell (e.g. Hold Person) will essentially always just attack — it never voluntarily casts the control spell, even when locking down a dangerous enemy would clearly be the stronger tactical choice.
Confirmed real, not theoretical
Proven by combinedFeatureParity.test.ts (milestone 13): a test scenario with both a cantrip and a control spell available had to be built with a melee-only cantrip to force the caster toward the spell at all, matching a workaround already used in runEncounter.spellcasting.test.ts (milestone 11). With any ranged attack option present, the control spell goes unused.
Suggested follow-up
Extend selectAction.ts's heuristics with a "worth it" evaluation for single-target save-based control effects — e.g. prefer a control spell against a target that's clearly the bigger threat (highest damage output, or the one currently threatening a low-HP ally), similar in spirit to the existing Int-driven-targeting and wasted-resource-avoidance heuristics already documented there.
This affects both the animated single-run viewer and Monte Carlo win-rate accuracy for any scenario with a support/control caster.
- Dominant language
- TypeScript
- Stars
- 0
- Forks
- 0
- Avg merge
- 1d 18h
- Merged PRs (30d)
- 2
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from RamonGebben/kernel-dm-toolbox
-
enhancement
Difficulty 5/5 Over a week Newbie friendliness 35/100
-
enhancement
Difficulty 5/5 Over a week Newbie friendliness 38/100
-
enhancement
Difficulty 5/5 Over a week Newbie friendliness 42/100
-
enhancement
Difficulty 4/5 3-5 days Newbie friendliness 66/100
-
enhancement
Difficulty 4/5 3-5 days Newbie friendliness 45/100
All issues in RamonGebben/kernel-dm-toolbox
Similar issues
-
clawsweeper:fix-shape-clear clawsweeper:queueable-fix clawsweeper:source-repro impact:ux-friction issue-rating: 🦞 diamond lobster no-stale P3
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
-
community first-timers-only good first issue hacktoberfest help wanted low hanging fruit up-for-grabs
Difficulty 1/5 Under an hour Newbie friendliness 76/100
-
code-quality refactoring
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
github/gh-aw-firewall#8816 ·
-
integration:quickjs org:external priority:backlog topic:code-interpreter topic:middleware type:feature
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
langchain-ai/deepagents#6450 ·
-
Difficulty 1/5 Under an hour Newbie friendliness 88/100
vercel/react-tweet#225 ·