The map — how the pieces connect

Fires on PreToolUse `*` (project rel + user). Blocks action tools (Edit/Write/Bash/etc.) until the agent prints `COHERENCE_CHECK`; read/search/oracle tools always pass

What it links to

• fires ask
• mentions calibrate
• blocks until coherence-check
• reads what it produces coherence-check
• fires decompose
• fires governer
• mentions jonathan-check
• mentions modal-inspect
• mentions pre-tool-governer-gate.sh
• mentions predict-required-skills
• mentions respond-to-direct-question
• mentions session-end-persist.sh
• reads what it produces session-start-coherence-gate-ab.sh
• fires speak-human
• mentions stop-complete-agentic-task.sh
• mentions subagent-start-governer.sh
• reads what it produces user-prompt-enrich.sh
• mentions reflect
• mentions infuse
• mentions smc

What links to it

• alignment-doctor reads what it produces this
• flywheel-consultant reads what it produces this
• instinct-harvest reads what it produces this
• pre-tool-obsidian-session-stamp.sh mentions this
• stop-evidence-gate.sh mentions this
• stop-slash-command-gate.sh mentions this
• orchestrator reads what it produces this

Fires on UserPromptSubmit (project, rel). The big one: scores the turn via governer API, arms the coherence gate file, tracks requested slash commands, injects coherence questions + verification contract + scope reflection into context

What it links to

• fires predict-required-skills
• runs smc
• runs smc-frustration-detector
• mentions align
• fires calibrate
• fires coherence-check
• mentions commit
• sends to decompose
• fires governer
• fires how-to-talk-like-the-founder
• reads what it produces reflect
• fires speak-human
• sends to verification-contracts

What links to it

• agent-debug reads what it produces this
• declare-scope reads what it produces this
• pre-tool-coherence-gate.sh reads what it produces this
• pre-tool-governer-gate.sh reads what it produces this
• seed mentions this
• stop-slash-command-gate.sh reads what it produces this
• alignment-doctor reads what it produces this
• ask reads what it produces this
• flywheel-consultant reads what it produces this
• instinct-harvest reads what it produces this

Use BEFORE asking the founder any clarifying question. Replaces naked questions with a SILA-gated decision flow — synthesize the context the question emerges from, predict the likely answer with a confidence number, score the question using IX-SILA (the canonical scoring system that's already run for the session), then route based on criticality. Trivial questions get answered by the agent and announced. Medium questions get an agent_find pass before reasking. High-criticality questions get the full briefing template and wait for greenlight. Communications use /how-to-talk-like-the-founder voice — no jargon, no menus, conveyed as a peer with advanced cognition speaking to another.

What it links to

• mentions governer
• mentions governer2
• must come before how-to-talk-like-the-founder
• mentions leverage-predictor
• reads what it produces user-prompt-enrich.sh

What links to it

• calibrate sends to this
• on-agent-getting-stuck sends to this
• on-turn-end sends to this
• post-tool-skill-invocation-annotation.sh mentions this
• pre-tool-askuserquestion-gate.sh blocks until this
• pre-tool-coherence-gate.sh fires this
• rapid-fire-testing sends to this
• session-start-coherence-gate-ab.sh fires this
• stop-on-turn-end-reminder.sh fires this

Re-ground to the standing-instructions file Core Protocol. Invoke when an agent is drifting from its core operating contract — communicating too technically/densely, skipping coherence checks, not reflecting intent before acting, or losing track of the alignment-first principle. Re-injects the first-principles operating contract from the standing-instructions file so the agent remembers what it's actually supposed to be doing. Especially important at end of substantive turns where agents lose protocol awareness amid context bloat.

What it links to

• sends to ask
• sends to expand-perspective
• runs how-to-talk-like-the-founder
• sends to reflect
• mentions coherence-check

What links to it

• on-turn-end sends to this
• pre-tool-coherence-gate.sh mentions this
• stop-on-turn-end-reminder.sh fires this
• stop-predict-required-skills2.sh fires this
• user-prompt-enrich.sh fires this

Per-message alignment ritual. Run at the top of every reply to confirm intent alignment with the founder before acting. Prints COHERENCE_CHECK header, surfaces intent, calibrates communication style, tracks pending scope, and determines whether to run /governer.

What it links to

• sends to decompose
• sends to governer
• runs how-to-talk-like-the-founder

What links to it

• on-complete-all-queued-tasks runs this
• pre-tool-coherence-gate.sh blocks until this
• pre-tool-coherence-gate.sh reads what it produces this
• session-end-instinct-harvest.sh reads what it produces this
• user-prompt-enrich.sh fires this
• on-agent-getting-stuck mentions this
• calibrate mentions this

SMC alignment injection — detects agent drift and injects targeted founder corrections. Use /smc to check status, verify telemetry, or troubleshoot.

What it links to

• reads what it produces jonathan-check
• runs smc-consumption-prompt
• mentions pre-commit-smc-prompt-guard.sh
• reads what it produces governer
• reads what it produces smc-frustration-detector

What links to it

• user-prompt-enrich.sh runs this
• pre-tool-coherence-gate.sh mentions this

Mine INSTINCTS (not facts) from founder-confirmed sanity checks and coherence checks in session transcripts, and maintain a per-domain instinct ledger that agents read before entering a situation. General tier — domain-agnostic. A domain layer (e.g. /instinct-harvest-seo) supplies the ledger path, the situation taxonomy, the session filter and extra seed sources, then invokes this skill. Use when seeding a domain's instinct ledger, when harvesting a session's confirmed checks, when generating the ledger index, or when checking whether the loop is healthy.

What it links to

• runs how-to-talk-like-the-founder
• reads what it produces sanity-check
• reads what it produces pre-tool-coherence-gate.sh
• reads what it produces user-prompt-enrich.sh
• reads what it produces session-start-verbatim-rehydrate.sh

What links to it

• session-end-instinct-harvest.sh runs this

Invoked directly via /rapid-fire-testing when the founder names an idea and wants the agent to race by the most direct path possible to one piece of tangible evidence proving whether the idea can actually produce the desirable outcome. Open-source-first, no stopping for permission, no perfecting prompts, no human-in-the-loop until first real artifact lands.

What it links to

• sends to ask
• mentions feature-completion-ux-test
• sends to governer
• runs how-to-talk-like-the-founder
• runs reflect

Fires on SessionStart (user). Picks coherence-prompt variant (control vs per-item-96-escalation) for hook #4

What it links to

• fires ask
• fires decompose
• fires governer
• fires speak-human

What links to it

• pre-tool-coherence-gate.sh reads what it produces this

Fires on SessionEnd (user). Detaches a node job that harvests SANITY_CHECK / COHERENCE_CHECK blocks + founder replies into the instinct ledger

What it links to

• reads what it produces coherence-check
• runs instinct-harvest
• reads what it produces reflect
• reads what it produces sanity-check

Detect frustration spikes in user messages and fire telemetry alerts

What it links to

• mentions smc-consumption-prompt

What links to it

• genome-experiment mentions this
• user-prompt-enrich.sh runs this
• smc reads what it produces this

Print a diagnostic overview showing what hooks fired, what got injected into context, what's silent, and what's shaping agent behavior — before every response. Use when the founder wants to see what's actually happening behind the scenes.

What it links to

• reads what it produces stop-complete-agentic-task.sh
• reads what it produces user-prompt-enrich.sh

Check whether institutional context is being injected into this turn. Reads the most recent hook output file, greps for the injection block, prints the result with an 8-line evidence extract.

What it links to

• reads what it produces pre-tool-coherence-gate.sh
• reads what it produces user-prompt-enrich.sh

What links to it

• pre-tool-coherence-gate.sh mentions this

Wire and protect the SMC consumption prompt that prefixes all alignment injections

What links to it

• smc runs this
• smc-frustration-detector mentions this

Fires on Stop (user). Blocks the stop if the human asked for `/commands` the agent never invoked

What it links to

• mentions pre-tool-coherence-gate.sh
• reads what it produces user-prompt-enrich.sh

Fires on PreToolUse `AskUserQuestion` (user). Blocks the AskUserQuestion popup; routes the agent to `/ask` skill instead

What it links to

• blocks until ask

Fires on PreToolUse obsidian write tools (project rel). Rewrites Obsidian MCP note payloads to stamp the Claude session ID into frontmatter

What it links to

• mentions pre-tool-coherence-gate.sh

Explain how agent_find works and troubleshoot its results. Use when asked how to query IX institutional knowledge.

What it links to

• reads what it produces knowledge-atom-extraction

What links to it

• decompose-nested-intent-from-transcript mentions this
• ultraplan2 runs this
• workflow-design runs this

Read cached research and plan docs from past sessions by ID. Use when agent_find returns an artifact_id.

What links to it

• hook-extract-on-stop.sh fires this
• research sends to this

What links to it

• agentic-find reads what it produces this

Fires on PostToolUse `agent_find` (user). Drops `.resolved` marker so gate #12 stands down

What links to it

• pre-tool-intent-search-gate.sh reads what it produces this
• session-start-intent-search-gate-arm-notice.sh reads what it produces this

What it links to

• reads what it produces post-tool-detect-agent-find.sh

What links to it

• session-start-intent-search-gate-arm-notice.sh reads what it produces this

Fires on SessionStart (user). Tells the agent to announce "Infusion of Institutional Context is Active" when gate #12 is armed

What it links to

• reads what it produces pre-tool-intent-search-gate.sh
• reads what it produces post-tool-detect-agent-find.sh

Fires on SubagentStop (user). Backgrounds artifact extraction from the subagent transcript

What it links to

• fires agentic-artifact-cache

Decompose a task into testable UX intent statements and post to compaction. Use at the start of every non-trivial task.

What it links to

• runs governer
• reads what it produces user-prompt-enrich.sh
• runs intent
• runs get-session-id.sh
• runs intent-db

What links to it

• decompose-nested-intent-from-transcript mentions this
• flywheel-consultant reads what it produces this
• gate sends to this
• gate runs this
• governer2 mentions this
• init runs this
• pipeline-scope-guardian runs this
• predict-required-skills sends to this
• session-start-alignment-autoload.sh fires this
• task-orchestrator runs this

Automatic task orchestration — ensures every agent task gets the right pipeline steps (/align, /governer, /decompose, /declare-scope, /reflect, /consume, /complete-agentic-task) without the founder having to ask. Predicts which steps are needed based on task characteristics, prints them, and spawns guardian agents.

What it links to

• runs align
• runs commit
• runs complete-agentic-task
• runs consume
• runs declare-scope
• runs decompose
• runs governer
• runs jonathan-check2
• runs predict-required-skills
• runs reflect
• runs speak-human
• sends to verify
• sends to mempalace-coaching-workflow

Real-time scope listener + skill router. Spawns a persistent team where a listener agent watches conversation for actionables, shows its interpretation, and routes to the right existing skill. Use when starting any substantive session.

What it links to

• runs agentic-init
• sends to align
• sends to declare-scope
• runs declare-scope
• sends to decompose
• runs decompose
• runs execute
• sends to founder-voice-email-writing
• sends to governer
• runs governer
• runs jonathan-check2
• runs speak-human

Expert on self-improving agent flywheels — systems where session data gets extracted into structured oracles that steer future sessions better. The IX alignment flywheel is the first working specimen of this pattern. Use this skill to understand how it works, audit signal quality, extend it to new data sources, diagnose degradation, OR to abstract the pattern and build a flywheel for a completely different domain. This is a case study AND a template.

What it links to

• reads what it produces align
• reads what it produces communicate-what-you-finished
• reads what it produces declare-scope
• reads what it produces decompose
• reads what it produces governer
• reads what it produces insight
• reads what it produces jonathan-check2
• reads what it produces reflect
• reads what it produces pre-tool-coherence-gate.sh
• reads what it produces user-prompt-enrich.sh

What links to it

• jonathan-check3 reads what it produces this

Use when about to compose copy that will live on a Next AI Labs / IX Coach / personal portfolio public surface — vehicle pages, intent pages, dedicated seed pages, audience-facing sections, CTA framings, or any prose a real reader will encounter as the organization's voice. The skill exists because this kind of work fails at the substrate level when an agent treats it as a composition task, and it succeeds when treated as an articulation-of-existing-intent task that selects and arranges from real founder source material. Read this skill end to end before producing any copy.

What it links to

• runs align
• mentions founder-voice-email-writing
• must come after how-to-talk-like-the-founder
• mentions how-to-write-inline-customer-messages
• sends to nested-intent2
• sends to oracles
• runs reflect
• mentions skill-file-create-pre-check-self-healing-protocol

What links to it

• founder-voice-email-writing mentions this
• how-to-talk-like-the-founder mentions this
• speak-human sends to this

Session-persistent adversarial execution team. Creates a team on first call that stays alive for the entire session — scope guardian, compliance auditor, and self-healing log accumulate context across all task batches. Dispatches implementers with independent cross-verification, adversarial disproof, and oracle alignment checking. Use when you have a confirmed task list and need verified implementation.

What it links to

• must come after align
• reads what it produces decompose
• must come after decompose
• runs governer
• runs jonathan-check2
• mentions workflow-factory
• sends to pipeline-adversarial-design

What links to it

• execute2 mentions this
• gate runs this
• ux-failure-decomposition must come after this

Take a specific transcript and produce an intent map of the client's system — restructured to lead with the critical opportunities and critical leverage, in the founder's exact voice, with every relevant verbatim quote integrated and noise removed. The artifact is the highest-fidelity signal about the client's intent — their actual words on the most substantive opportunities, with all critical context that makes those opportunities lucid, clear, and born. Use when a transcript needs to become a customer-reviewable map that aligns all parties around the leverage points the founder would have surfaced if writing it himself.

What it links to

• mentions agentic-find
• runs align
• mentions declare-scope
• mentions decompose
• mentions gap
• mentions intent-architect
• runs reflect
• mentions speak-human

Isolate a specific measurable gap in the current scope, predict what's next, and design a plan in human terms that communicates the gap boundaries and what will fix it. Use when you need to zoom in from a large scope to one concrete, actionable piece.

What it links to

• reads what it produces align
• runs speak-human

What links to it

• decompose-nested-intent-from-transcript mentions this
• on-turn-end sends to this
• session-start-alignment-autoload.sh fires this
• stop-on-turn-end-reminder.sh fires this
• superpowers-writing-plans-v2 runs this
• ultraplan2 runs this

Produce a nested intent map for a multi-layered initiative — top-level goal, child goals, sub-goals, cross-cutting intents — where each node carries an intent paragraph in founder voice with enough precision that a subagent reading just that node can act at 100% fidelity. Use when the human asks for nested intent on an initiative with multiple sub-goals, when a plan needs to be turned into a context-managed handoff package for subagents, or when an existing outline needs intent paragraphs added at each layer of nesting.

What it links to

• sends to align
• sends to decompose
• runs how-to-talk-like-the-founder
• sends to reflect
• runs speak-human

What links to it

• aligned-build runs this
• skill-file-create-pre-check-self-healing-protocol sends to this
• writing-aligned-user-facing-words sends to this

Deep-research planning workflow. Takes any scope and iterates through reflect → align → gap-identification → research-agent dispatch → oracle consultation → repeat until 95% confidence → outputs a full intent-nested decomposed plan, gated by jonathan-check2, written to Obsidian and opened. Use when a scope is complex, ambiguous, or high-leverage and requires multi-source research before committing to a plan.

What it links to

• runs agentic-find
• runs align
• runs decompose
• runs gap
• runs insight
• runs intent-nested
• runs jonathan-check2
• runs reflect

Team-based workflow that produces human-readable journey documentation grounded in institutional knowledge with governing agents auditing every section. Use when documenting any user-facing flow, conversion funnel, or system behavior that needs founder validation. Supersedes v1 for new work — v1 remains available.

What it links to

• runs align
• runs hallucination
• runs insight
• runs speak-human

What links to it

• anthropic-proof-its-fixed runs this
• workflow-extract-from-session runs this

Extract a reproducible workflow from the current session's methodology. Use when a session produced a breakthrough result and you want to capture the exact steps, corrections, and quality gates into a permanent skill that any future agent can invoke. Invoke when the founder says 'capture this', 'make this reproducible', 'I don't want to lose this', or 'extract this into a workflow'.

What it links to

• runs intent-journey-documentation-v2
• runs align
• runs insight
• runs reflect
• mentions speak-human
• runs workflow-design

God-mode founder oracle — queries the IX god-mode notebook loaded with all extracted session data (coherence checks, reflect outputs, align outputs, skill sequences, compaction intents, scope declarations). Predicts what the founder would say AND what the next step should be. Loops autonomously until the oracle signals scope is complete. Everything printed observably.

What it links to

• reads what it produces flywheel-consultant
• replaces jonathan-check2
• reads what it produces reflect
• reads what it produces align

What links to it

• on-agent-getting-stuck sends to this

Read this before creating or updating any skill file. The founder's articulation, in his own words, of how to encode intent with precision so that the encoded artifact actually produces capacity in future agents who consume it rather than reducing their capability to apply intent to context. Includes a self-healing protocol — when the founder is dissatisfied with a skill, the agent reviews the calibration with a third-party agent, prints its critique, responds, and fixes, narrating the process.

What it links to

• sends to nested-intent2
• sends to reflect

What links to it

• founder-voice-email-writing mentions this
• how-to-talk-like-the-founder mentions this
• writing-aligned-user-facing-words mentions this

Alignment-coaxed clone of superpowers:writing-plans (v1). Additively enforces IX Coach's alignment protocol: Obsidian filing, open-on-create, required frontmatter, ascendant-chain confirmation, probability-qualified claims, speak-human tone, §0.4 handoff contract, kill-switch section, additive-feedback default, parent-agent self-enforcing loop awareness. v1 behavior preserved byte-accurate in PART A.

What it links to

• runs gap
• must come after how-to-talk-like-the-founder
• runs on-talking-about-code
• runs reflect
• runs speak-human

Diagnostic protocol invoked when an agent stated or acted on something false as if it were fact.

What it links to

• mentions extract-coaching-principle

What links to it

• fix-hallucinated-code-cleanup-source runs this
• intent-journey-documentation-v2 runs this
• validate-load-bearing-claims-against-reality mentions this

Surface unfinished work from the current Claude session transcript and print it to the founder in his own voice for approval. Sonnet does the analysis. Nothing is written to any database — the founder decides what happens to each item.

What it links to

• runs how-to-talk-like-the-founder
• replaces incomplete
• runs on-talking-about-code
• runs speak-human

Fires on SessionStart (user). Injects the alignment-trigger text block at session start

What it links to

• fires declare-scope
• fires gap
• fires governer

What links to it

• post-compaction-reload.sh mentions this

Fires on SubagentStart (user). Scores the subagent's task via API and injects a hard "state the target UX before writing code" requirement + agent_find directive

What it links to

• mentions governer
• fires speak-human
• fires verification-gate

What links to it

• pre-tool-coherence-gate.sh mentions this

Agent-invoked completion gate. When an agent believes its work is done, this skill forces it to collect evidence — load the page, grab screenshots, run verification commands from the seed's success criteria. The agent invokes this itself when it's ready. No hooks, no interrupts.

What links to it

• align must come before this
• decompose must come before this
• initiative-playground must come before this

Reusable pre-code playground for any major, user-impacting initiative. Before touching real code, model the whole thing as wikilinked Markdown files (runtime vs build-harness), anchor the founder's verbatim intent, decompose quotes to plans, make every decision observable, run A/B tests in chat, log every correction back into the model, and refine the response on evidence until it's right. Only then build. Use when the founder says "explore this potential / I want to build X / let's sandbox this / can we get a better response before we touch code" for anything that would impact users.

What it links to

• must come before complete-seed
• must come before governer

What links to it

• align must come before this

MANDATORY discipline for any communication that references code — to the founder, in comments, in docs, in proposals, in reports. A code reference is worthless-to-negative unless it is fused with the exact literal conditions (numbers, states, inputs the system actually checks) AND the precise user journey that makes the statement true. Force-read whenever you reference code, and especially on triggers like "what exact ux", "i don't understand", "what are the conditions". Read this BEFORE writing the code reference, not after.

What links to it

• incomplete2 runs this
• superpowers-writing-plans-v2 runs this

Trace a user journey through the codebase — tell the story of what a person experiences, step by step, stopping at the first break. Use when investigating bugs, auditing flows, explaining code changes, or grounding proposals in reality.

What it links to

• runs align
• runs speak-human

Identify code changes based on hallucinated intent, trace the source material that caused the hallucination, fix the code, then

What it links to

• runs hallucination

Curate and map human intent into validated seed documents with nested structure, certainty scores, and predict-and-validate communication

What links to it

• decompose-nested-intent-from-transcript mentions this

Consume the nested intent map for a product, derive every way each intent could fail in actual use, decompose into concrete validation tasks that return with deterministic evidence. The skill that asks "does this actually work for real?" — not "does the code exist?

What it links to

• must come after execute

Final-pass auditor that verifies every code citation in a research/synthesis answer against actual files BEFORE publishing. Use at the end of any code-grounded research task. Catches the retrieval-augmented confabulation pattern where agents invent specific file/line citations that feel grounded but don't exist.

What links to it

• subagent-start-governer.sh fires this

Outcome-first verification protocol. Forces the agent to define what the OUTCOME looks like in measurable terms BEFORE gathering any evidence. Blocks the failure pattern of substituting mechanism-evidence (hook is configured, file exists, command runs) for outcome-evidence (the thing the human wants to be true is actually happening). Use whenever asked to check if something is working.

What links to it

• task-orchestrator sends to this

Queries the Agent Skill Pattern Oracle (NotebookLM) with the current scope/intent and returns a MANDATORY skill sequence grounded in historical patterns. Each predicted skill MUST become a TaskCreate item. Integrates with /align, /governer, /decompose, /declare-scope. Caches result — only re-predicts on scope change.

What it links to

• sends to admin-tooling-templates-full-stack
• sends to agentic-development
• sends to align
• sends to coaching-eval-system
• sends to coaching-skills-system
• sends to coaching-telemetry
• sends to code-principles-how-to-code
• sends to commit
• sends to consume
• sends to declare-scope
• sends to decompose
• sends to fix-algo-bug-determine-if-auto-fix-ok
• sends to forecasting-and-metrics
• sends to governer
• sends to jonathan-check2
• sends to reflect
• sends to skill-creator
• sends to validate-against-live-data
• sends to verification-contracts

What links to it

• governer sends to this
• governer2 sends to this
• pre-tool-coherence-gate.sh mentions this
• user-prompt-enrich.sh fires this
• task-orchestrator runs this
• predict-required-skills2 replaces this
• predict-required-skills2 mentions this

EXPERIMENTAL clone of /governer with FORCED workflow stages, A/B-tested oracle enrichment, and full observability. Each required skill returned by the oracle becomes a TaskCreate item agents cannot skip. Every invocation announces variant + WHEN it runs in /speak-human form, prints full oracle input/output transcript, and logs any forced-stage violations. Behind GOVERNER2_ORACLE_ENRICH and GOVERNER2_FORCED_STAGES flags — both default off. Created by Lattice 2026-04-23 under L-SEED-001 + L-G2-001.

What it links to

• sends to commit
• runs consume
• mentions declare-scope
• sends to discoverability-audit-learnings-findable-by-agents
• sends to governer
• reads what it produces governer
• sends to predict-required-skills
• runs sanity-check
• runs speak-human
• mentions stop-evidence-gate.sh

What links to it

• aligned-build runs this
• ask mentions this
• execute2 mentions this
• founder-watchable-canonical sends to this

Use when a top-level agent must define outcomes, route work across analysis, research, and implementation layers, preserve

What it links to

• runs execute2
• runs compact-agentic-session
• sends to governer
• must come after governer
• runs intent-db
• runs speak-human
• runs session-titles-name
• reads what it produces pre-tool-coherence-gate.sh

What links to it

• commit sends to this
• init runs this
• multi-tier-audit-pipeline mentions this
• workflow-design runs this

Fires on Stop + SubagentStop (user). Blocks the stop: if the last message claims something works/is fixed/deployed and there is no verification evidence in the transcript, forces the agent to go verify

What it links to

• reads what it produces governer
• mentions pre-tool-coherence-gate.sh
• mentions whats-left
• reads what it produces verification-contracts
• blocks until verification-before-completion

What links to it

• agent-outcome-summary must come after this
• governer2 mentions this
• loop mentions this

Task-specific verification contracts generated by governer at task start, enforced by the Stop hook at task end.

What it links to

• runs intent
• replaces seed
• reads what it produces governer

What links to it

• pipeline-consumption-handoff sends to this
• predict-required-skills sends to this
• user-prompt-enrich.sh sends to this
• stop-evidence-gate.sh reads what it produces this

Load the SMC seed document into session memory for alignment work and founder-intent context.

What it links to

• mentions jonathan-check
• mentions user-prompt-enrich.sh

What links to it

• close sends to this
• deploy-alignment-guardian must come after this
• seed-company mentions this
• proposal-schema mentions this
• verification-contracts replaces this

Use when performing any multi-step audit, triage, or alignment task where future agents may repeat the same work.

What it links to

• sends to principle-alignment-scanner-workflow
• mentions admin-tooling-templates-full-stack
• mentions agentic-tooling-registry
• mentions how-to-submit-and-track-proposals
• mentions run-discoverability-audit
• mentions system-log-triage

Dispatch a single subagent on a task with full alignment discipline baked in — ascendant-context read, falsifiable verification contract, RECOVERABLE_DRAFTS handoff, and parent self-enforcing loop. Use when you (parent agent) want to delegate a discrete task to a subagent and get back work that you can independently verify with runnable commands instead of trusting self-assessment. Portable: works for any task in any project. Validated 2026-04-23 by T13-LBL-001 substrate test (9/9 V1-V9 PASS) within the auto-calibrating-agent-os initiative; this skill extracts that pattern as a reusable command.

What it links to

• must come after dispatch-seed
• mentions execute
• sends to governer
• must come after governer
• mentions governer2

What links to it

• orchestrator runs this

Bootstrap a session by activating orchestrator mode, confirming governer calibration, and arming declare-scope — all in one command. Use at the start of any session instead of running /orchestrator, /governer, and /declare-scope separately.

What it links to

• runs declare-scope
• runs governer
• runs orchestrator
• mentions speak-human

What links to it

• extract-wip-from-past-sessions runs this

Automated 6-phase audit that maps the standing-instructions file coding rules to scanner detectors, finds violations, and fixes them safely.

What it links to

• runs fix-algo-bug-determine-if-auto-fix-ok
• sends to fix-error-logs
• sends to get-error-logs

What links to it

• track sends to this

What it links to

• reads what it produces governer
• reads what it produces user-prompt-enrich.sh
• blocks until governer

What links to it

• pre-tool-coherence-gate.sh mentions this

Deploy the Alignment Guardian agent for the current project. Use when governer scores a task >= 50, or when the human requests persistent alignment monitoring.

What it links to

• reads what it produces align
• must come after seed
• reads what it produces governer

Systematically eliminate a bug class by identifying the root cause, scoping all vulnerable call sites, building a

What links to it

• predict-required-skills sends to this
• principle-alignment-scanner-workflow runs this

Use when you are about to report or act on a load-bearing conclusion — especially one a governer scored 90+ — that you reached by reading code, docs, or any single source of inference. This skill is the discernment that turns operating-certainty into validated truth by measuring the actual reality your claim describes. Triggers — high-leverage conclusion, "the code shows", "the DB does X", "this resolves to", "the gate fires", reporting a finding you have not measured against the thing it is a claim ABOUT, any 90+ governer task before the completion claim.

What it links to

• must come after governer
• mentions hallucination
• mentions verification-before-completion

What links to it

• governer sends to this

Fires on SessionStart (user). Re-injects most recent governer score after compaction recovery

What it links to

• fires governer
• mentions session-start-alignment-autoload.sh

Load the IX Coach founding mission — why IX Coach exists, the Silver Grail thesis, and the alignment-first philosophy — in the founder's actual voice. Use when you need customer-facing mission language, when drafting communications that connect someone to the "why", or when an agent needs the founding story for retention or outreach.

What it links to

• mentions seed
• mentions silver-grail-advisor

Create new skills, improve existing skills, and measure skill performance.

What links to it

• predict-required-skills sends to this

Brief on open, recent, or historical sessions in natural language. Use when asking what's up or what's incomplete.

What it links to

• must come before governer

Rename unnamed Claude Code sessions with intent-rich titles using batched haiku subagents. Dispatches as background orchestrator task.

What links to it

• orchestrator runs this

Confirms that work matches intent — before starting, after completing, before committing, or when reviewing a plan or diff. U

What it links to

• runs how-to-talk-like-the-founder
• runs speak-human
• sends to intent-journey-documentation

What links to it

• batch runs this
• batch-whats-next must come after this
• batch-whats-next reads what it produces this
• commit reads what it produces this
• founder-watchable-canonical runs this
• founder-watchable-canonical must come before this
• governer runs this
• governer2 runs this
• instinct-harvest reads what it produces this
• pipeline-cross-verify runs this
• session-end-instinct-harvest.sh reads what it produces this

What it links to

• sends to audit-pre-commit
• mentions batch
• sends to defensive-error-audit
• sends to fix-console-log-hotpath
• sends to how-to-use-feature-flags
• sends to staged-release-management

What links to it

• predict-required-skills sends to this

What it links to

• sends to defensive-error-audit
• sends to fix-console-log-hotpath
• sends to how-to-use-feature-flags
• sends to staged-release-management

What links to it

• batch runs this
• code-principles-how-to-code sends to this

Review uncommitted files, group them by concern, and walk through each batch with UX impact analysis before committing. Default to the Obsidian dashboard pattern — one reviewable note per concern in a one-click approve/reject/request-evidence surface — for any multi-concern sweep.

What it links to

• runs audit-pre-commit
• runs commit
• sends to governer
• runs sanity-check

What links to it

• batch-whats-next reads what it produces this
• code-principles-how-to-code mentions this

Present finished, already-verified work — a batch ready to commit, a delegated fix that came back, a re-proposal after a gap was closed — as a set of per-item cards the founder can say yes/no/hold to, each one framed as the intent that must be true if the work is aligned, never as a description of what the code now does.

What it links to

• reads what it produces batch
• mentions how-to-talk-like-the-founder
• must come after sanity-check
• reads what it produces sanity-check
• sends to work-status-report

The second report — after the substantive report or answer, show every piece of work grouped by how far it has actually got (proposed / work in progress / done not committed / done, committed, merged to staging), each item written as one conversational paragraph carrying what it is, the intent read back for the founder to validate or reject, what happens next, and what he will then have. Use whenever reporting status, what is left, or what is actionable.

What it links to

• runs how-to-talk-like-the-founder

What links to it

• batch-whats-next sends to this

Sweep a range of past Claude Code sessions with a fleet of Sonnet agents (one per session) to surface implied-but-unfinished work the founder wanted done, filter to a target initiative (e.g. the gptcoach2 launch), and convert each finding into a priority-prefixed, self-contained Obsidian proposal sorted into work-state triage folders. Use when the founder wants to mine recent or older sessions for dropped actionables before a launch, or says "extract WIP", "sweep the last sessions", "what did I leave unfinished", or "run the WIP sweep on the next N sessions".

What it links to

• runs align
• runs canonical-ui-patterns
• runs commit
• runs governer
• runs how-to-submit-and-track-proposals
• runs init
• runs jonathan-check2
• runs on-turn-end
• runs pipeline-consumption-handoff
• mentions predict-required-skills2
• runs resolve-proposals
• runs speak-human
• runs task-get-one-important
• runs update-the-founder

On task completion, surface the work product to the founder as a list of UX changes — one sentence per change in "when X then a user would see Y because Z but D" form, with emojis flagging which were verified in Chrome. Use after finishing any code change, UX fix, or feature work that produced observable user-facing outcomes. Pairs with /update-the-founder when you also need a synthesized status check-in. Triggers: task complete, work done, finished, ship summary, what did you change, list ux changes, completion summary, what got built, hand off.

What it links to

• runs consume
• runs predict-next-action
• sends to agent-outcome-summary
• sends to update-the-founder

What links to it

• agent-outcome-summary replaces this
• agentic-init sends to this
• flywheel-consultant reads what it produces this
• pipeline-consumption-handoff must come after this
• workflow-design runs this

Run at the end of any substantive turn that produced a work product with a governer score >30. Routes you to the right consumption skill so the founder can see, understand, and act on what you did in seconds — never reading pages of output to figure out what you were working on, which part you just touched, or where to look at it. Skip for pure conversation/brainstorm with no work product.

What it links to

• sends to ask
• sends to calibrate
• sends to consume
• sends to gap
• runs propose
• sends to update-the-founder
• reads what it produces governer

What links to it

• extract-wip-from-past-sessions runs this
• stop-on-turn-end-reminder.sh fires this

Founder-voice status check-in after every substantive turn — 1-7 sentences that instantly clarify what you're working on, what just happened, what the intent was, what's predicted next, with quick single-letter commands attached. Replaces verbose agent narration with synthesized signal the founder can scan in 5 seconds instead of reading paragraphs. Use after finishing any substantive task, when the founder says START, when handing off mid-work, when you sense the founder needs to observe state for human-in-the-loop review, or when about to ask a clarifying question and want to surface state first. Triggers: status update, what are you working on, where are you, founder update, check in, update me, what's the status, surface state, START, observability, human in the loop, what just happened, where do things stand.

What it links to

• sends to consume
• runs how-to-talk-like-the-founder
• sends to jonathan-check2
• mentions reflect

What links to it

• communicate-what-you-finished sends to this
• extract-wip-from-past-sessions runs this
• on-turn-end sends to this
• stop-on-turn-end-reminder.sh fires this

Preserve the founder's authentic voice in all outreach, emails, and client-facing copy. Founder's verbatim words with grammar fixes only — every AI "improvement" is a degradation.

What it links to

• mentions skill-file-create-pre-check-self-healing-protocol
• mentions writing-aligned-user-facing-words

What links to it

• extract-completions-for-log sends to this
• gate sends to this
• how-to-talk-like-the-founder must come before this
• how-to-talk-like-the-founder sends to this
• writing-aligned-user-facing-words mentions this

Produce a consumption document for every completed pipeline item before handing off to the human. Use when work is finished and the human needs to know exactly what was built, how to open it, and what to look for. NEVER hand over a URL without verifying it first.

What it links to

• must come after communicate-what-you-finished
• must come after consume
• runs speak-human
• sends to verification-contracts

What links to it

• extract-wip-from-past-sessions runs this
• reflect must come before this
• workflow-design runs this

Stage 10 of the observable pipeline. Generates fresh high-leverage ideas to feed the pipeline queue when it runs low. Queries multiple sources (jonathan-check2 oracle, leverage-predictor, predict-next-action, agent_find for errors/complaints, compaction incomplete items), scores each candidate via governer methodology, formats as pipeline-ready items, and appends to PIPELINE-TRACKER.md after deduplicating against existing runs/.

What it links to

• runs governer
• mentions how-to-submit-and-track-proposals
• runs jonathan-check2
• runs leverage-predictor
• mentions next-priority-tasks
• runs predict-next-action

What links to it

• workflow-design runs this

Fires on Stop (user). Non-blocking nudge to run `/on-turn-end` when governer score > 40 or unknown

What it links to

• fires ask
• fires calibrate
• fires consume
• fires gap
• fires on-turn-end
• fires propose
• fires update-the-founder

Generate a trustworthy, human-legible outcome report after completing dispatched work. Each item translates code changes into UX-testable statements with verification evidence grounding. The founder reads this in 30 seconds and knows what happened.

What it links to

• replaces communicate-what-you-finished
• must come after stop-evidence-gate.sh

What links to it

• communicate-what-you-finished sends to this
• dispatch-seed runs this
• founder-watchable-canonical sends to this

Fix approved proposals and close the loop — claim, fix, commit with Resolves-Proposal tags, mark implemented. U

What it links to

• sends to intent-db

What links to it

• commit2 sends to this
• extract-wip-from-past-sessions runs this

Pipeline from intent alignment to agent dispatch. Runs /align until confirmed, creates observable seed in Obsidian, generates plan with docs-first approach, then dispatches. The full lifecycle from "I want this done" to "an agent is working on it.

What it links to

• runs agent-outcome-summary
• runs align

What links to it

• execute2 must come after this

What links to it

• communicate-what-you-finished runs this
• pipeline-idea-generator runs this
• predict-next-action2 mentions this

Use when the founder asks for a daily progress log, end-of-week roundup, "what did I get done", Obsidian dashboard entry, or any summary block describing recent work that he will read. Produces signal-trigger blocks written in his own voice — verb-led headings, short prose paragraphs, inline obsidian wikilinks, plain date stamp. Read the `how-to-talk-like-the-founder` skill voice samples BEFORE drafting any block — non-negotiable.

What it links to

• sends to founder-voice-email-writing
• runs how-to-talk-like-the-founder

Read, write, and search the Founder Knowledge Store — personal insights, sales lessons, relationship patterns, strategic f

What links to it

• how-to-talk-like-the-founder sends to this

Surface the single most critical proposable task from the current conversation — derives gap between user intent and reality, fills the full SessionCompaction incompleteItem proposal schema, and presents it in human terms.

What links to it

• extract-wip-from-past-sessions runs this