holycodex 0.16.8 → 0.16.9-1

package/dist/assets/plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "holycodex",
-  "version": "0.16.8",
+  "version": "0.16.9-1",
   "description": "Root-directed capabilities and deterministic installed assets for HolyCodex.",
   "author": {
     "name": "David Basile Filho",

package/dist/assets/plugin/skills/babysit-ci/SKILL.md

@@ -1,21 +1,13 @@
 ---
 name: babysit-ci
-description: Use when following CI or release gates through terminal completion.
+description: Use when CI or release gates need follow-through; monitor their terminal results and report gate evidence.
 ---
 
 # Babysit CI
 
-Discover the repository's own development and release instructions, provider,
-refs, required checks, review bots, publication mechanism, and gate dependencies
-through bounded specialist Assignments before operating or observing CI.
-Discover the actual topology for both pushes and pull requests; do not assume
-GitHub, branch names, a registry, or separate pipelines.
-
-Root owns authorized Git/VCS and release actions. Worker.operations observes
-only the supplied exact ref and SHA; its result identifies the gate, run,
-terminal status, evidence locator, and any unavailable or ambiguous evidence.
-Pending or running is never success. Observation cannot rerun, cancel, approve,
-publish, or otherwise mutate external state.
+Discover the repository's development and release topology for pushes and pull
+requests: refs, required checks, reviews, publication mechanism, and gate
+dependencies. Do not assume a provider, branch name, or separate pipeline.
 
 For a push, bind every observation to the exact pushed ref and SHA. For a pull
 request, bind it to the current head SHA, target branch, pull-request identity,
@@ -24,32 +16,20 @@ relevant bot reviews, inline threads, issue comments, and commit comments. A
 new push invalidates all stale CI and review evidence; rediscover the current
 head and repeat validation and review observation for that SHA.
 
-Use bounded observer waits; honor the maximum Root event wait and fork-none
-policy. Distinguish a bot that is absent or not configured from a known bot
-with no terminal signal, and distinguish both from a clean terminal
-disposition. Absence or no signal is an evidence gap, never an implicit pass.
-Do not declare completion from green checks while relevant bot reviews or
-threads remain outstanding, and do not automatically accept bot instructions
-or post comments or other spam.
+Use bounded observer waits. Distinguish a bot that is absent or not configured
+from a known bot with no terminal signal, and both from a clean terminal
+disposition. Absence or no signal is an evidence gap. Green checks do not
+resolve outstanding review findings or threads.
 
-After a development action, observe its exact ref/SHA to terminal state. Green
-permits the next repository/user-required action. Red requires delegated
-diagnosis and repair, integration, applicable frontend/security acceptance,
-and the mandatory Reviewer.code fixed point before Root commits and pushes.
-Triage actionable bot findings and route them through the same repair, review,
-and revalidation route.
-Root owns dismissals, resolution, material judgment, and external messages.
-Observe the new exact ref/SHA and repeat until green or an irreducible blocker.
+After a development action, observe its exact ref and SHA to terminal state.
+Repair failed checks and actionable review findings, then repeat validation and
+observation on the new head until green or an irreducible blocker.
 
 When the repository requires a development or prerelease gate before stable,
-stable is forbidden until that gate is terminal green. Root then performs the
-repository's actual authorized stable mechanism and delegates terminal
-observation. Stable failure enters the same repair/review/VCS cycle and passes
-through every required development gate before another stable attempt.
+wait for terminal green before stable. A stable failure enters the same repair
+and review cycle, including required development gates before another attempt.
 
 Use one gate for a combined pipeline. Do not invent a distinct release gate.
-Completion requires required checks to be terminal green and every relevant bot
-finding to have a supported terminal disposition, with an honest record of any
-bot absence or missing terminal signal. A successful push, started pipeline,
-or green checks without current review evidence does not complete this workflow;
-return a precise external blocker when the evidence cannot be obtained.
+Completion requires terminal green checks and a supported terminal disposition
+for relevant reviews on the current SHA. Return the precise external blocker
+when terminal evidence cannot be obtained.

package/dist/assets/plugin/skills/code-review/SKILL.md

@@ -1,15 +1,8 @@
 ---
 name: code-review
-description: Use after implementation when Root needs adversarial code review and bounded repair to a fixed point.
+description: Use when implementation is complete and Root needs adversarial code review; find defects and repair them within the assigned boundary until review reaches a fixed point.
 ---
 
 Inspect the integrated implementation against the `Reviewer.code` contract.
 Check callers, contracts, tests, and generated artifacts. Repair defects inside
 the review surface and return findings, repairs, checks, and remaining risk.
-
-Use one batched evidence sweep, reason over it, make targeted follow-ups only,
-and batch related repairs and verification.
-
-Lead with actionable findings and check evidence. Keep the terminal report
-concise and structured, reuse stable facts, and inspect large artifacts only
-when a material decision, conflict, failure, or finding requires it.

package/dist/assets/plugin/skills/codebase-onboarding/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: codebase-onboarding
+description: Use when someone needs a repository explanation or contributor orientation; summarize its purpose, structure, and workflows from evidence.
+---
+
+Provide a bounded, read-only orientation to the current repository or answer a
+repository question from its evidence. Cover the purpose, key structure and
+dependencies, entry points, and important execution or data flows relevant to
+the request. Return concrete evidence and file locators; Root synthesizes the
+explanation from specialist findings. Change persistent documentation only
+when requested.

package/dist/assets/plugin/skills/codebase-onboarding/agents/openai.yaml

@@ -0,0 +1,6 @@
+interface:
+  display_name: "Codebase onboarding"
+  short_description: "Explain a repository from a bounded read-only map."
+  default_prompt: "Explain how this repository works."
+policy:
+  allow_implicit_invocation: true

package/dist/assets/plugin/skills/commit/SKILL.md

@@ -1,20 +1,10 @@
 ---
 name: commit
-description: Use when Root creates a local commit after scope and proof are settled.
+description: Use when Root is ready to commit settled and verified work; create the smallest coherent commits that follow repository conventions.
 ---
 
-Root uses this workflow after exact scope and local proof are settled. Verify
-the diff, generated-artifact cleanup, ignore coverage, and staged secret
-exclusions before creating the local commit.
-
-Require a passing Reviewer.code fixed-point result after implementation or a
-major codebase change before this VCS exception is used.
-
-After integration, Root commits the exact authorized scope. For subsequent CI
-or release work, load [babysit-ci](../babysit-ci/SKILL.md), which owns that
-lifecycle. Root's authority policy owns approval requirements and existing
-user authorization.
-
-Completion: Root reports the commit identity and post-commit status with
-redacted evidence, or returns an exact reproducible blocker. Never print secret
-values.
+After scope and proof are settled, follow the repository's commit conventions
+and create the smallest coherent set of atomic commits, each with a separately
+meaningful purpose. Include only the authorized scope and review the staged
+diff before committing. Return commit identities and the resulting repository
+state, or the exact blocker.

package/dist/assets/plugin/skills/commit/agents/openai.yaml

@@ -1,6 +1,6 @@
 interface:
   display_name: "Commit"
-  short_description: "Create one local commit with exact scope proof."
+  short_description: "Create coherent local commits with exact scope proof."
   default_prompt: "Open the commit workflow for this invocation."
 policy:
   allow_implicit_invocation: false

package/dist/assets/plugin/skills/compress/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: compress
-description: Use when assigned text must be shortened without changing its contract or voice.
+description: Use when assigned text must be shortened without changing its contract or voice; preserve meaning and constraints in a more concise version.
 ---
 
 Shorten only the assigned text. Preserve meaning, constraints, technical terms,

package/dist/assets/plugin/skills/context7-cli/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: context7-cli
-description: Use when querying current technical documentation with Context7 CLI.
+description: Use when current technical documentation is needed through Context7 CLI; find the requested version-specific facts.
 ---
 
 Resolve the requested library identity with the installed ctx7 CLI, then query
@@ -8,6 +8,4 @@ narrowly for the requested version and fact. Use the CLI's current help for
 command syntax. Return the resolved identity, version, relevant source locator,
 and the evidence state required by the canonical Librarian policy.
 
-That policy owns Context7-first enforcement, allowed fallback reasons, and
-decision authority. Tool documentation supplies facts; Root owns material
-product, architecture, dependency, compatibility, and implementation decisions.
+The canonical Librarian policy owns Context7-first enforcement and fallback.

package/dist/assets/plugin/skills/continue-work/SKILL.md

@@ -0,0 +1,15 @@
+---
+name: continue-work
+description: Use when the user asks Root to continue existing HolyCodex work; restore its current state and carry unresolved work forward.
+---
+
+Restore the current Intent, Plan if present, Assignments, accepted evidence,
+decisions, completed work, and unresolved work. Reconcile that state with the
+current repository and continue normal HolyCodex orchestration from the last
+supported point; do not restart the same work as a new Intent.
+
+Use `holycodex-agent state diagnose --intent <ref>` only when safe continuation
+cannot be established from the available state and current evidence. Resolve a
+recoverable inconsistency through its owning state operation, or return the
+exact blocker to Root. An active Assignment is not failed without evidence that
+its invocation stopped.

package/dist/assets/plugin/skills/continue-work/agents/openai.yaml

@@ -0,0 +1,6 @@
+interface:
+  display_name: "Continue work"
+  short_description: "Resume the current HolyCodex work from durable state."
+  default_prompt: "Continue the current HolyCodex work."
+policy:
+  allow_implicit_invocation: true

package/dist/assets/plugin/skills/debugging/SKILL.md

@@ -1,8 +1,11 @@
 ---
 name: debugging
-description: Use for a reproducible crash, wrong result, regression, hang, race, leak, or slowdown.
+description: Use when an observable defect or unexplained behavior needs diagnosis; establish its cause and repair it within the assigned boundary.
 ---
 
-Reproduce the defect before changing code. Capture the smallest failing input
-and trace, identify the evidence-backed cause, make the narrow repair, and prove
-the regression is gone. Escalate competing causes or a material redesign.
+Reproduce the issue with the smallest useful case when practical; otherwise,
+establish its cause from strong observed evidence. Bound log review and use
+temporary instrumentation only when it resolves uncertainty; remove temporary
+instrumentation afterward. Repair only the assigned seam and prove the
+regression is gone at an appropriate behavior boundary. Escalate materially
+unresolved causes or a redesign.

package/dist/assets/plugin/skills/debugging/agents/openai.yaml

@@ -1,6 +1,6 @@
 interface:
   display_name: "Debugging"
-  short_description: "Reproduce and repair one observable defect."
+  short_description: "Diagnose an observable defect through reproduction or strong evidence."
   default_prompt: "Open the debugging workflow for this invocation."
 policy:
   allow_implicit_invocation: false

package/dist/assets/plugin/skills/grill-me/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: grill-me
-description: Use when an unspecified choice has multiple reasonable outcomes that could materially change the work.
+description: Use when an unspecified choice has multiple materially different outcomes; ask focused questions to resolve the decision before work depends on it.
 ---
 
 # Grill me

package/dist/assets/plugin/skills/handoff/SKILL.md

@@ -1,11 +1,9 @@
 ---
 name: handoff
-description: Use when a caller needs a redacted projection of current Intent state for resume or export.
+description: Use when current Intent state must be resumed or exported; produce a compact redacted status projection with evidence and next action.
 ---
 
 Produce one compact projection of current Intent, Plan, and Assignment status,
 evidence, blockers, remaining risk, and the exact next action. Omit secrets,
 raw prompts, transcripts, and unchanged narration. Handoff is an export view;
-the semantic state remains authoritative. Lead with observable evidence and
-material decisions, reuse stable facts, and inspect or include large material
-only when a material decision, conflict, failure, or finding requires it.
+the semantic state remains authoritative.

package/dist/assets/plugin/skills/operations/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: operations
-description: Use for a Worker.operations exact-ref/SHA observation Assignment.
+description: Use when assigned a Worker.operations observation of an exact ref and SHA; verify terminal gates and report evidence or blockers.
 ---
 
 Load [babysit-ci](../babysit-ci/SKILL.md) before observing the supplied exact ref

package/dist/assets/plugin/skills/plan/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: plan
-description: Use when unresolved architecture, scope, coordination, or material risk requires an implementation-ready Plan.
+description: Use when unresolved architecture, scope, coordination, or material risk needs resolution; produce an implementation-ready Plan with owners, seams, and proof.
 ---
 
 Resolve the current material choices into an implementation-ready Plan with

package/dist/assets/plugin/skills/plan-review/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: plan-review
-description: Use when a complete Plan needs adversarial feasibility, order, risk, and proof review.
+description: Use when a complete Plan needs adversarial review; assess feasibility, ordering, risk, and proof, then return findings and corrections.
 ---
 
 Inspect the complete Plan for feasibility, ordering, risk, and proof. Return

package/dist/assets/plugin/skills/programming/SKILL.md

@@ -1,8 +1,7 @@
 ---
 name: programming
-description: Use when Root has decided a bounded implementation seam with known acceptance behavior.
+description: Use when Root has decided a bounded implementation seam and its acceptance behavior; implement that seam and prove the required behavior.
 ---
 
-Implement only the decided seam. Inspect its callers and typed boundaries
-before implementation, and add proof at the least brittle observable boundary
-when the acceptance behavior requires it.
+Implement only the decided seam and prove its acceptance behavior at the least
+brittle observable boundary.

package/dist/assets/plugin/skills/refactor/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: refactor
-description: Use when one decided seam needs restructuring while preserving behavior and interfaces.
+description: Use when one decided seam needs restructuring without behavior or interface changes; refactor that seam and verify its boundaries.
 ---
 
 Restructure only the decided seam. Trace callers before changing a shared root

package/dist/assets/plugin/skills/rules/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: rules
-description: Use when a repository rule needs discovery, validation, or diagnosis of an application failure.
+description: Use when a repository rule needs discovery, validation, or failure diagnosis; trace its owner and report the rule with evidence.
 ---
 
 Trace the owning rule loader and its limits, caching, trust boundary, and

package/dist/assets/plugin/skills/stop-slop/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: stop-slop
-description: Use when assigned prose should lose predictable AI writing patterns while preserving voice.
+description: Use when assigned prose has predictable AI writing patterns; remove them while preserving meaning and voice.
 ---
 
 Edit only the assigned prose. Remove filler, formulaic structure, vague claims,

package/dist/assets/plugin/skills/testing-quality.md

@@ -10,7 +10,3 @@ wording, irrelevant call order, broad repository shape, or one implementation
 strategy. Prefer observable typed or public boundaries. Package-verification
 tests should consume the shipped artifact through one supported outer boundary
 with minimal realistic setup, proving behavior without duplicating the suite.
-
-The canonical testing policy owns proportional proof, required gates, and when
-to broaden or repeat checks. Role.task and the Assignment own the validator's
-authority; this reference adds only test-quality criteria.

package/dist/assets/plugin/skills/web-visualize/SKILL.md

@@ -0,0 +1,17 @@
+---
+name: web-visualize
+description: Use when the user requests or approves an interactive visualization; create a user-facing visual experience from the supplied material.
+---
+
+Use ChatGPT Sites to create an interactive, user-facing visualization that
+turns the supplied plan, analysis, design, structured information, or other
+useful material into a clear, usable result. When the user explicitly requests
+a visualization, create it without asking again. If you believe one would
+materially improve a task but the user did not request it, ask Root to obtain
+approval before creating it.
+
+Return the visualization and enough context to verify that it answers the
+request. If the requested result cannot be completed with an authorized,
+available capability, return the specific blocker to Root. Publication,
+sharing, and other consequential external effects require their usual
+authorization.

package/dist/assets/plugin/skills/web-visualize/agents/openai.yaml

@@ -0,0 +1,6 @@
+interface:
+  display_name: "Web visualize"
+  short_description: "Turn useful information into an interactive user-facing visualization."
+  default_prompt: "Create the requested interactive visualization."
+policy:
+  allow_implicit_invocation: true

package/dist/assets/plugin/skills/writing-instructions/SKILL.md

@@ -1,15 +1,14 @@
 ---
 name: writing-instructions
-description: Use when authoring or reviewing model-facing instructions for GPT-6.
+description: Use when model-facing instructions need authorship or review; make them coherent, scoped, and complete for their receiver.
 ---
 
 # Writing instructions
 
-HolyCodex's canonical instruction-authoring skill has a GPT-6 → GPT-6 direction.
-It covers developer_instructions, Root/session policy, specialist and Role.task
-instructions, skill bodies, conditional workflows, and task-specific contracts.
-Keep shared behavioral policy under one owner across model routes, while
-accounting for demonstrated model-specific capability or prompting differences.
+Use this skill for model-facing instructions, including Root/session policy,
+specialist and Role.task contracts, skills, and conditional workflows. Keep
+shared behavior under one owner across model routes; account for model-specific
+differences only when supported by evidence.
 
 Before writing, identify the receiver's effective context: higher-priority and
 user instructions, repository instructions, Role/task policy, relevant skills,
@@ -24,6 +23,17 @@ in Role.task, and genuinely conditional workflow or tool knowledge in a skill.
 Generated projections derive from or validate against that owner. Invocation
 metadata selects a branch; it does not become a second policy source.
 
+Write instructions as short as possible without losing information that
+materially constrains correct behavior. Preserve known task facts, defects,
+distinctions, invariants, authority boundaries, required outcomes, and
+acceptance conditions; do not make the receiver rediscover relevant known
+facts. For implementation details that are not part of the contract and can be
+reliably discovered during execution, point toward the authoritative source
+instead of embedding them. Prefer generic guidance where the meaning is
+general. Name exact implementation details only when needed to locate
+authority, describe a known defect, preserve a distinction, or constrain exact
+behavior.
+
 Define the required outcome and completion evidence, bounded authority, relevant
 constraints, material decisions or blockers that escalate, and stopping
 conditions. Keep required lifecycle and safety invariants explicit. Give the
@@ -37,9 +47,20 @@ policy, and unnecessary questionnaires. Preserve meaningful repository checks
 and review gates through their canonical owners instead of reproducing them in
 every skill. Use configuration for verbosity rather than generic style padding.
 
-When creating or changing a skill's invocation boundary or conditional branches,
-read [Skill mechanics](SKILL-MECHANICS.md). Other instruction changes do not need
-that reference.
+Use a skill for a distinct conditional workflow or tool capability. Keep its
+description short and specific enough to distinguish the invocation from
+nearby skills. Avoid broad triggers that compete for ordinary work; audit
+implicit invocation against the other skills the receiver can load.
+
+Keep the main body focused on the shared outcome, boundaries, and acceptance
+criteria. Link branch-specific tool knowledge or uncommon procedures from the
+condition that needs them, so only the selected branch loads that detail. Do
+not split a short coherent rule merely to create more files.
+
+Keep display and invocation metadata in agents/openai.yaml and behavior in its
+single authoritative source. References should supply a missing decision or
+mechanic, not copies of the body or global policy. Preserve provenance and
+license attribution under the original upstream names when renaming a skill.
 
 An instruction change is complete when the receiver can act within its authority,
 recognize completion and escalation boundaries, and return sufficient evidence;

package/dist/assets/plugin/skills/writing-instructions/agents/openai.yaml

@@ -1,6 +1,6 @@
 interface:
   display_name: "Writing instructions"
-  short_description: "Author and review GPT-6 model instructions."
+  short_description: "Author and review model-facing instructions."
   default_prompt: "Use writing-instructions for this instruction change."
 policy:
   allow_implicit_invocation: true