@zalom/plastic 2.0.0-alpha.26 → 2.0.0-alpha.27

package/scripts/read-config

@@ -1,5 +1,5 @@
 #!/usr/bin/env ruby
-# Usage: read-config <key> [--default VALUE] [--project PATH]
+# Usage: read-config <key> [--default VALUE] [--project PATH] [--harness claude|codex]
 # Resolves a config value: project -> global -> built-in defaults.
 # Nested keys use dot notation: agent.type, architect.style
 # Returns scalar values as strings, hash/array values as JSON.
@@ -46,6 +46,7 @@ def parse_args(argv)
   key = nil
   default_val = nil
   project_dir = nil
+  harness = nil
 
   i = 0
   while i < argv.length
@@ -58,13 +59,17 @@ def parse_args(argv)
       abort "Error: --project requires a value" unless argv[i + 1]
       project_dir = argv[i + 1]
       i += 2
+    when "--harness"
+      abort "Error: --harness requires a value" unless argv[i + 1]
+      harness = argv[i + 1]
+      i += 2
     else
       key = argv[i]
       i += 1
     end
   end
 
-  [key, default_val, project_dir]
+  [key, default_val, project_dir, harness]
 end
 
 def load_yaml(path)
@@ -119,10 +124,10 @@ if ARGV.include?("--migrate")
   exit 0
 end
 
-key, explicit_default, project_dir = parse_args(ARGV)
+key, explicit_default, project_dir, harness = parse_args(ARGV)
 
 if key.nil? || key.empty?
-  $stderr.puts "Usage: read-config <key> [--default VALUE] [--project PATH]"
+  $stderr.puts "Usage: read-config <key> [--default VALUE] [--project PATH] [--harness claude|codex]"
   exit 1
 end
 
@@ -135,9 +140,42 @@ if project_dir
   project_config = load_yaml(project_config_path)
 end
 
-value = dig_key(project_config, key)
-value = dig_key(global_config, key) if value.nil?
+def canonical_harness(value)
+  return nil if value.nil?
+  return "claude" if %w[claude claude-code].include?(value.to_s)
+  return "codex" if value.to_s == "codex"
+
+  abort "Error: --harness must be claude or codex"
+end
+
+def harness_agent_value(config, key, harness)
+  match = key.match(/\Aagents\.(models|efforts)\.([^.]+)\z/)
+  return nil unless match
+
+  section, agent = match.captures
+  if section == "models"
+    AgentModels.models_section(config, harness: harness)[agent]
+  else
+    AgentModels.efforts_section(config, harness)[agent]
+  end
+end
+
+harness = canonical_harness(harness)
+if harness
+  value = harness_agent_value(project_config, key, harness)
+  value = harness_agent_value(global_config, key, harness) if value.nil?
+else
+  value = dig_key(project_config, key)
+  value = dig_key(global_config, key) if value.nil?
+end
 value = explicit_default if value.nil? && explicit_default
+if value.nil? && harness && (match = key.match(/\Aagents\.(models|efforts)\.([^.]+)\z/))
+  value = if match[1] == "models"
+            AgentModels.shipped_model_for(match[2], harness: harness)
+          else
+            AgentModels::DEFAULT_EFFORT
+          end
+end
 value = dig_key(DEFAULTS, key) if value.nil?
 
 puts format_value(value)

package/skills/agent-advisor/SKILL.md

@@ -6,7 +6,7 @@ description: >-
   failed attempts, or ranking several plausible options. Use when the user asks for
   a second opinion, a hard design decision, an architecture review, help breaking a
   deadlock, or says "ask the advisor". Also sets which advisor is the default when
-  asked ("make Fable my advisor", "switch my advisor", "use the real advisor").
+  asked ("make Primary Advisor my default", "switch my advisor", "use Secondary Advisor").
 user-invocable: true
 ---
 
@@ -15,12 +15,10 @@ user-invocable: true
 Plastic ships two consultation agents, never dispatched by the auto pipeline, summoned
 only when you decide the reasoning is worth buying:
 
-- **`plastic-advisor`** ("the real advisor"): the frontier model itself, expensive,
-  billed through usage credits. Spawn it for a few rounds on the hardest problem, then
-  close the session.
-- **`plastic-faux-advisor`** ("the imitation advisor"): an ordinary model carrying the
-  same reasoning discipline inline in its own body, so it reasons the same disciplined
-  way at a fraction of the cost. The cheaper default.
+- **Primary Advisor** (`plastic-primary-advisor`): Fable at medium effort. Use it for
+  normal consultation.
+- **Secondary Advisor** (`plastic-secondary-advisor`): Fable at high effort. Use it only
+  as an explicit escalation for harder or higher-risk reasoning.
 
 ## When to consult (and when not to)
 
@@ -33,25 +31,23 @@ when the ordering decides where you spend the next day.
 Never buy a consultation for: anything a tool can answer (search, reading code, running
 tests, documentation), writing code at volume, confirming a decision you already made,
 style or naming a linter would settle, or anything reversible and cheap you have not
-tried first. The full buy/never-buy list, the effort table, and the entry test live in
+  tried first. The full buy/never-buy list, answer-shape table, and entry test live in
 `references/advisor-protocol.md`; read it before writing a brief for the first time in
 a session.
 
 ## Routing: which advisor answers
 
 1. Read the harness-scoped config: `advisor.claude.default`, the only advisor routing key
-   the installer writes. If unset, use `plastic-advisor`, the shipped default.
-2. If the user names which advisor they want ("ask the real one", "use Fable", "ask the
-   cheap one"), honor that directly and dispatch `plastic-advisor` (the real advisor) or
-   `plastic-faux-advisor` (the cheaper imitation) accordingly, overriding step 1 for this
-   consultation only.
+   the installer writes. If unset, use `plastic-primary-advisor`, the shipped default.
+2. If the user explicitly asks for Primary Advisor or Secondary Advisor, honor that choice for
+   this consultation. Otherwise, dispatch the configured agent from step 1.
 3. If `advisor.enabled` reads `false`, neither advisor agent nor this skill is
    installed; this step should not be reachable, but if it is, tell the user the
    advisor is disabled and point at "Setting the default" below.
 4. Dispatch the resolved agent with a brief built per `references/advisor-protocol.md`
    section 4 (natural prose, the block is a completeness check, not a form to fill).
-   State EFFORT and the answer shape explicitly; classify low and prove your way up,
-   never open high "to be safe".
+   State the answer shape explicitly. Primary uses medium effort and Secondary uses high;
+   only an explicit owner config override changes either value.
 5. Consume the answer per the protocol's section 5: run the Operating Manual's
    five-question self-test on the advisor's plan before executing it. Advice is input,
    not authority; the plan is the advisor's, the outcome is yours.
@@ -66,27 +62,23 @@ ruby ~/.plastic/scripts/read-config advisor.claude.default --project <repo>
 
 ## Setting the default advisor
 
-When asked to change the default ("make Fable my advisor", "switch my advisor", "use
-the cheaper one by default"), present the two options in plain language and write the
+When asked to change the default, present the two options in plain language and write the
 choice:
 
-- **Faux Fable** (`plastic-faux-advisor`, recommended): an ordinary model carrying the
-  frontier reasoning instructions. Much cheaper, available on any plan, reasons in the
-  same disciplined way.
-- **Fable 5** (`plastic-advisor`): the frontier model itself. The strongest reasoning
-  available, billed through usage credits, so summon it for a few rounds and close it.
+- **Primary Advisor** (`plastic-primary-advisor`, recommended): Fable at medium effort.
+- **Secondary Advisor** (`plastic-secondary-advisor`): Fable at high effort.
 
 These are the same two options the installer offers at install and update time. Write
 the choice to `advisor.claude.default` in the global `~/.plastic/config.yml` (or the
 project's `.plastic_store/config.yml` when the user scopes the change to one project):
-read the file as YAML, set `advisor.claude.default` to the agent name (`plastic-advisor`
-or `plastic-faux-advisor`, never a model name or nickname), and write it back. Confirm
+read the file as YAML, set `advisor.claude.default` to the agent name (`plastic-primary-advisor`
+or `plastic-secondary-advisor`, never a model name or nickname), and write it back. Confirm
 the new default back to the user in one line.
 
 ## References
 
 - `references/advisor-protocol.md`: the full shipped Advisor Protocol (what to buy,
-  effort and answer shape, the entry test, how to write a brief that earns its cost, the
+  answer shape, the entry test, how to write a brief that earns its cost, the
   answer contract, session economics, anti-patterns). Read it before the first
   consultation in a session; the second consultation in the same advisor thread costs a
   fraction of the first, so keep follow-ups on one thread rather than opening a new one.

package/skills/agent-advisor/references/advisor-protocol.md

@@ -4,13 +4,13 @@
 
 *Adapted for Plastic (intent 185): this is the shipped reference copy the
 `plastic-agent-advisor` skill teaches from. Two named agents carry it,
-`plastic-advisor` (the real advisor) and `plastic-faux-advisor` (the cheaper
-imitation, the same discipline at a fraction of the cost); EFFORT and the
-answer shape below shape the brief and the answer on whichever one you
+`plastic-primary-advisor` (Primary Advisor) and `plastic-secondary-advisor`
+(Secondary Advisor, the explicit escalation); the
+answer shape below shapes the brief and the answer on whichever one you
 dispatch, never which file to pick. The
 `plastic-agent-advisor` skill reads harness-scoped config
-(`advisor.claude.default`, falling back to `secondary`, then to
-`plastic-faux-advisor`) to route your consultation automatically; name a
+(`advisor.claude.default`, falling back to `plastic-primary-advisor`) to route your
+consultation automatically; name a
 specific advisor in your request to override it. A frontier model rewards a
 single, well-formed one-shot brief more than a back-and-forth, so front-load
 everything section 3 below asks for before you send. The rest of this
@@ -53,17 +53,18 @@ You own the work and the outcome. Fable owns the hardest thinking, on demand.
 
 ---
 
-## 2. Effort and answer shape
+## 2. Answer shape
 
 Classify every consultation before sending it. Default to the smallest shape and
-prove your way up. Never open at a higher effort "to be safe": effort follows the cost of being
-wrong, not the importance the task feels like it has.
+prove your way up. Primary Advisor ships at medium reasoning effort. Secondary Advisor ships
+at high reasoning effort on Claude Code and Codex. An explicit harness-scoped config override
+may change one agent's effort.
 
 | Shape | Coding | Business | Research | Effort | Brief size | Expected return |
 |------|--------|----------|----------|--------|------------|-----------------|
-| **Verdict** | Verdict on one step; choose between two named implementations | Pick between two options you already compared (vendor, pricing point) | Judge whether one source or result is trustworthy enough to build on | `low` | Under 300 words | Verdict, one paragraph of reasoning, the single biggest risk |
-| **Plan** | Plan a feature inside one system; review a full plan for holes; design one interface; rank root causes | Positioning or pricing decision from a compiled evidence pack; review a proposal before sending it | Design a research plan for a bounded question; rank competing explanations of the data you gathered | `medium`, or `high` if an irreversible step is inside | Up to one page | Decision, numbered plan with per-step checks, risk map |
-| **Architecture** | Cross-system architecture; migration with one-way doors; deadlock after two failed attempts; security-critical design | Build-vs-buy, market entry, or any commitment measured in months; strategy where reversal is expensive | Synthesis across many sources where the conclusion drives a large bet; contested questions with conflicting evidence | `xhigh`; `max` only when being wrong means data loss, a broken contract, or weeks of rework | Full evidence brief | Decision, plan, risk register, kill criteria, list of what could not be verified |
+| **Verdict** | Verdict on one step; choose between two named implementations | Pick between two options you already compared (vendor, pricing point) | Judge whether one source or result is trustworthy enough to build on | `medium` | Under 300 words | Verdict, one paragraph of reasoning, the single biggest risk |
+| **Plan** | Plan a feature inside one system; review a full plan for holes; design one interface; rank root causes | Positioning or pricing decision from a compiled evidence pack; review a proposal before sending it | Design a research plan for a bounded question; rank competing explanations of the data you gathered | `medium` | Up to one page | Decision, numbered plan with per-step checks, risk map |
+| **Architecture** | Cross-system architecture; migration with one-way doors; deadlock after two failed attempts; security-critical design | Build-vs-buy, market entry, or any commitment measured in months; strategy where reversal is expensive | Synthesis across many sources where the conclusion drives a large bet; contested questions with conflicting evidence | `medium` | Full evidence brief | Decision, plan, risk register, kill criteria, list of what could not be verified |
 
 **Escalate one shape when any of these holds:**
 
@@ -124,11 +125,8 @@ What actually raises the quality of my answer, in order of impact:
    under the Operating Manual." Then I write steps you can run at your best:
    each step with its own check, its own trap named, and its own
    stop-and-return trigger. A plan without that is half a plan.
-6. **Know what effort buys.** At `low` I stress-test your candidate and give a
-   verdict. At `medium`/`high` I generate rival solutions and compare them. At
-   `xhigh`/`max` I build the strongest case for every rival and then try to
-   break my own winner before you ever see it. Buy the depth the failure cost
-   justifies, nothing more.
+6. **Use the medium default well.** Ask for the smallest answer shape that can decide the
+   question, and spend the available reasoning on the load-bearing risk.
 7. **Come back on the same thread.** My context is cached inside a session.
    The second question in a thread costs a fraction of the first. A new session
    pays for your whole brief again.
@@ -142,7 +140,7 @@ briefing described above; use the block as your completeness check before
 sending. Fable must never need to explore.
 
 ```
-SHAPE: verdict | plan | architecture        EFFORT: low | medium | high | xhigh | max
+SHAPE: verdict | plan | architecture
 DOMAIN: coding | business | research
 GOAL: <target state in one sentence, and the decision this answer feeds>
 QUESTIONS:
@@ -229,8 +227,8 @@ reformat, then work with what you have.
   hear yes.
 - **Drip-feeding.** Splitting one decision across many small calls, paying
   session overhead each time.
-- **Prestige escalation.** Requesting `max` because the task feels important.
-  Effort follows failure cost, nothing else.
+- **Prestige escalation.** Changing the medium default because the task feels important,
+  without evidence that a harness-scoped override will improve the decision.
 - **Unbounded questions.** "Any thoughts on this approach?" invites an essay.
   Ask for a decision with named options.
 - **Silent adoption.** Pasting Fable's plan straight into execution without

package/skills/auto/SKILL.md

@@ -15,8 +15,8 @@ Announce: "Taking over intent [ID] - [name] for autonomous delivery."
 **Advisory (not a rule).** At auto-mode start, recommend once that the user run this
 orchestrating session on the best available thinking model (Fable, Opus, or whatever supersedes
 them); this is advice only, and dispatched agents keep their configured model, never resolving
-to Fable without an explicit `agents.models.<name>` config override. `plastic-advisor` and
-`plastic-faux-advisor` are consultation roles the user or this session summons deliberately;
+to Fable without an explicit `agents.models.<name>` config override. Primary Advisor and
+Secondary Advisor are consultation roles the user or this session summons deliberately;
 the auto pipeline never dispatches them.
 
 ## Precondition

package/skills/install/SKILL.md

@@ -94,16 +94,14 @@ Ask the user one feature question, interactive sessions only:
 
 If yes, ask which advisor is the default, exactly two choices:
 > "Which advisor should be the default?"
-> - **Faux Fable** (recommended): Opus 4.8 carrying the frontier reasoning
->   instructions. Much cheaper, available on any plan, reasons in the same
->   disciplined way. -> append `--advisor faux`
-> - **Fable 5**: the frontier model itself. The strongest reasoning available,
->   billed through usage credits, so summon it for a few rounds and close it. ->
->   append `--advisor real`
+> - **Primary Advisor** (recommended): Fable at medium effort for normal
+>   consultation. -> append `--advisor primary`
+> - **Secondary Advisor**: Fable at high effort for explicit escalation. ->
+>   append `--advisor secondary`
 
 Non-interactive sessions (no tty) skip the question entirely: the install ships with the
 shipped default, advisor enabled with no `--advisor` flag (the `plastic-agent-advisor`
-skill's own routing falls back to `plastic-faux-advisor` at consult time).
+skill's own routing falls back to `plastic-primary-advisor` at consult time).
 
 Update flow: pending config questions, including this one, are now announced
 generically by `plastic-update`'s Step 2, sourced from `config_asks.yml` - not

package/skills/update/SKILL.md

@@ -81,7 +81,7 @@ Once the user picks an option, run the printed `write-config` command for that
 option, for example:
 
 ```
-ruby ~/.plastic/scripts/write-config advisor.claude.default plastic-faux-advisor
+ruby ~/.plastic/scripts/write-config advisor.claude.default plastic-primary-advisor
 ```
 
 If they say "not now" / want to keep the default, run the printed dismissal

package/templates/config.yml

@@ -34,18 +34,25 @@ architect:
 #       plastic-enforcer: opus
 #     codex:
 #       plastic-executor: gpt-5.6-terra
+#   efforts:
+#     claude:
+#       plastic-executor: medium
+#     codex:
+#       plastic-executor: medium
+# Every shipped agent defaults to medium reasoning effort on both harnesses.
+# An agents.efforts.<harness>.<agent> entry is an explicit per-agent override.
 
-# The advisor (intent 185): two consultation agents, plastic-advisor (the real
-# advisor, expensive, model fable by default) and plastic-faux-advisor (the
-# cheaper imitation, model opus by default, carrying the same reasoning
-# discipline inline). advisor.enabled defaults to enabled; missing or
+# The advisors: Primary Advisor and Secondary Advisor. Both use Fable on Claude
+# Code and gpt-6-astra on Codex. Primary uses medium effort; Secondary uses high.
+# advisor.enabled defaults to enabled; missing or
 # unreadable counts as enabled (fail-open). The claude.default value is an
 # AGENT NAME, never a model name, so it can point at
 # a locally registered agent instead. Set at install time via --no-advisor /
 # --advisor, or uncomment here directly. Each agent's actual model is a plain
-# agents.models.claude.<name> override, same mechanism as any other agent.
+# agents.models.claude.<name> override, same mechanism as any other agent. Codex
+# generates both advisors with their role-specific effort.
 # advisor:
 #   enabled: false               # skip installing both advisor agents and the
 #                                 # agent-advisor skill entirely
 #   claude:
-#     default: plastic-advisor         # which agent the advisor skill routes to
+#     default: plastic-primary-advisor # which agent the advisor skill routes to

package/templates/node-research.md

@@ -3,9 +3,11 @@ node: r1
 kind: research
 files: []
 budget: 60000
+# Optional. When present, use one relative Markdown path under resources/.
+# report: resources/report--topic.md
 ---
 # r1 - <what to find out>
 
 ## Deposit
-<Where the finding lands: a resources/ file, an artifact URL, or a ruling
-recorded in the intent's ## Decisions.>
+<Where the finding lands: declare report: above for a resources/ Markdown file,
+or name an artifact URL or a ruling recorded in the intent's ## Decisions.>