TAOM

Root-cause native Bannerlord CTDs (AccessViolation in TaleWorlds.Native.dll) via Event Log offsets, offline disassembly, and live debugger forensics. No symbols needed.

Dispatch an independent Codex verification job directly via `codex exec` Bash call, then retrieve and present results

Inspect all changed files, group by logical concern, and guide atomic per-concern commits following TAOM's 50/72 rule and commit trailer convention.

Load a snapshot saved by /context-save and present it so this session can pick up where the last one left off without re-deriving decisions.

Launch parallel deep-dive agents to review completed work for quality, standards, compatibility, completeness, and cross-system data flow

Respond to a Bannerlord engine update (Steam force-bump or deliberate migration) — preserve the decompile baseline, regen, diff, re-verify bindings and creature data, control-battle.

Integrate a merge-ready branch into the trunk — fast-forward check, merge, regenerate backlinks, CHANGELOG, delete branch (local+remote), push with confirmation. TAOM trunk-based, not Git Flow.

Six-phase root-cause debugging for crashes, build failures, or "why is this broken" reports. Iron Law - no fixes without root cause. Auto-engages /freeze.

Compile source material from docs/raw/<topic>/ into a navigable wiki node at docs/research/<topic>.md. Use when source-material has accumulated and needs summarization with cross-refs to existing feature docs / RCAs / memory.

Autonomous doc-rot loop — fix one lint finding at a time, re-lint, commit on improvement; runs until clean or N stuck iterations.

Auto-detect what was built, write a Codex adversarial prompt, dispatch Codex directly via `codex exec`, then verify results when complete

Audit context window consumption across TAOM agents, skills, rules, MCP servers, and CLAUDE.md. Report token estimates, flag bloat, recommend trims.

Use before patching, overriding, or adapting a TaleWorlds type — `taom-src path <Type>` decompiles + caches the installed engine, returning an absolute path.

Author a culture's LOTRLOME armor items and swap troop equipment rosters via the generator + validation pipeline. Use when adding/revamping armor. Enforces the canonical-folder + cover-attribute rules.

Snapshot current working context (git state, in-flight tasks, key decisions, modified files) so a future session or post-compact resumption can pick up without losing progress. Pair with /context-restore.

Run the doc-health linter (tools/lint_docs.py) over docs/ and summarize dead links, stale version refs, orphan/missing feature docs. Use when auditing knowledge-base health or after large doc moves.

Use when asked to audit the repo for improvement opportunities (bugs, security, perf, tests, tech debt, game data), suggest direction, or write self-contained handoff plans.

Use when reviewing an external repo or article to adopt practices into TAOM — security-vet first, map novel vs duplicative, port (never install), review, commit your changes.

Use when auditing TAOM's Claude Code config for security issues — committed secrets, over-broad permissions, hook exfiltration, MCP risk, hidden-unicode injection. Runs tools/audit_claude_config.py.

Periodic quality audit of installed skills + agents. Quick scan (recently changed only) or full audit. Catches decay (broken cross-references, stale paths, bloated descriptions) before sessions silently degrade.

Propagate new player-facing text through TAOM's 12-language localization pipeline — wrap with {=KEY}, register the string, run the Claude translation tool, validate. Use after adding UI text or XSLT strings.

Use when cleaning AI-writing tells out of a commit body, CHANGELOG entry, issue/PR, or doc. Strips slop; keeps TAOM's em-dash/boldface house style.

Use when a creature's locomotion or mounted-rider animation looks wrong (pace/float/walk-in-place/idle) and needs refining or authoring in Blender before Kit-compile.

Author a rideable creature mount end-to-end (assets, Monster/action/usage XML, C# behavior tree, validation) following the elephant+spider-proven workflow. Warg parity is law.

Use when navigating or auditing the docs/ knowledge graph — explain a doc's links, find the path between two docs, or surface god nodes / bridges / orphans.

Orchestrate the mandatory completion sequence before merging a C# feature — /verify, /deep-review, fix findings, /review-codex, then close the issue and update docs + CHANGELOG.

Give TAOM lords lore-driven skill values and traits via the TAOM SkillSet system. Use when a canonical lord has wrong stats or a culture roster needs a balance pass.

Author or revamp a TAOM culture's armor set, troop tree, and recruitment wiring end-to-end. Use for new cultures or troop-tree revamps. Follows docs/ai-includes/new-culture-authoring.md.

Verify TAOM's Harmony patch, GameModel, and reflection bindings against the installed Bannerlord engine and refresh the committed API snapshot. Use after an engine update or patch change.

Assess whether a proposed change fits the current work context or represents scope creep

Structured 4-phase self-debug for failing agent runs (looping, drifting, burning tokens). Capture, diagnose, contained recovery, report. Complements /investigate (which is for code bugs); this skill is for harness/agent failures.

Hard-block Edit/Write outside a chosen directory. Use to scope-lock edits during a focused fix or refactor. Pair with /unfreeze to release.

Incrementally fix dotnet build errors with minimal diffs, one error at a time

Scaffold a new TAOM feature module with IoC, services, adapters, and tests following project conventions

Release the directory edit lock set by /freeze. Edit and Write are unrestricted again. No-op if /freeze was never set.

Create a GitHub issue for a feature, bug fix, or crash with all required TAOM sections

Scaffold a new Architecture Decision Record, auto-numbered from existing ADRs, with context pre-filled from recent git history and CHANGELOG.

Regression-safe cleanup of AI-generated bloat in C# code. Deletion-first, requires green tests before starting.

Run comprehensive build + test + git verification and produce a pass/fail report

Decompile and analyze TaleWorlds classes before implementing or fixing. Use when touching Harmony patches, adapters, GameModels, or investigating bugs.