Consilium — Approach Board
Purpose
Examine a problem, or an approach someone already picked, from several angles and several scopes at
once. The board answers four questions: what are we actually deciding, what approaches exist,
what does the one on the table foreclose, and what would a different framing buy.
It is not a review skill. Defects appear here only as
evidence that an approach is wrong — a bug
that is fixable inside a candidate is not this board's output. A diff goes to
.
Autonomous: run every step without asking the user, and present the report when done. Consilium
reads and reasons; it never modifies the thing it examines.
When to Use
- "how should we approach X", "what are the options", "what's the best way to", "explore approaches"
- "is this the right approach", "stress-test this plan", "think hard", "ultrathink"
- Before committing to a decision that is expensive to reverse — a data model, a public contract, a
dependency, an architectural direction
- ,
Not this skill:
| You have | Reach for |
|---|
| A diff, and you want it attacked for bugs and requirement gaps | |
| Comment noise, naming, convention drift in changed code | |
| One quick outside opinion, no board and no synthesis | |
| A back-and-forth about a design, not a verdict | |
Cost. This is the most expensive skill in the collection. Auto-selection keeps the board itself
at four or five seats, and verification adds
three more dispatches on the largest payload of the
run — so a typical run is seven or eight agents and a full board is nine. On a board of four or fewer
seats, run the
lens alone: with one critic there is nothing to corroborate, and
is the
lens that changes outcomes. Spend this skill on decisions that are expensive to reverse, not on
questions one seat could answer.
Focus Areas (optional )
| Focus | Board | Use case |
|---|
| (default) | Core + auto-selected | Let the board select its own optional seats |
| All six | Every angle, highest cost |
| Core + Librarius | Likely already solved somewhere |
| Core + Scrutator | Blast radius and what the choice forecloses |
| Core + Censor | Suspect overbuilding, or a simpler option skipped |
| Core + Scrutator + Censor | Big decision, no prior-art question |
The Seats
Three generate, three critique. The generators never see each other's output; the critics see the
assembled candidate set and nothing about who produced which candidate.
Core (always)
| Seat | Job | Phase |
|---|
| Novator | Proposes fundamentally different candidate approaches, each concrete enough to start | Diverge |
| Peregrinus | An agent outside this process, with none of this conversation's context — frames the problem cold | Diverge |
| Seneca | Attacks the framing and the load-bearing assumptions; checks the candidates are genuinely distinct | Converge |
Optional (auto-selected)
| Seat | Job | Launch when | Skip when |
|---|
| Librarius | Prior art — has this been solved, what do comparable systems and libraries already do | The problem sounds general, names a library or ecosystem, or looks like a well-trodden shape | Genuinely internal, domain-specific, or no external surface |
| Scrutator | Scope and blast radius — what each candidate touches, forecloses, and locks in; second-order effects | Wide reach, migrations, public contracts, data models, irreversible choices | Local, cheap to undo, contained in one module |
| Censor | Proportionality — cost to build and operate against the size of the problem; is a simpler candidate being skipped | New abstractions, multi-part machinery, anything that smells larger than the problem | Already minimal, or the cost is the point |
Selection happens twice, because the evidence arrives twice. Librarius is a generator, and its
trigger is a property of the frame, so decide it in Phase 1. Scrutator and Censor judge candidates
that do not exist until Phase 3 — deciding them from the frame is guessing, so decide them in Phase 3
against the assembled set. State each include/skip decision with a one-line reason at the point you
make it.
A board whose only critic is Seneca is a defective board. Seneca attacks the framing; nobody is
then looking at reach or at cost. If the assembled set contains any candidate that touches a contract
outside this codebase, reshapes stored data, or is expensive to leave, Scrutator runs. If any
candidate is materially larger than another that survives, Censor runs. Reaching Phase 4 with one
critic is allowed only when the candidates are genuinely small and cheap to undo, and the report must
say the board had one critic.
Minimum 3 (core only). Maximum 6. Typical 4–5 seats, plus verification.
Choosing Models and Depth
Stated as intent, since the roster changes and each host names its own models:
- Give each seat the most capable model the host offers. If that is the model running this skill,
take the next tier down — a seat on the orchestrator's own model shares whatever the orchestrator
already believes about this problem.
- Where the host lets you pick a model per seat, give each a different one. Same-model seats
differ only by sampling; same-role seats only by phrasing. Where it does not, run the default and
say so in the report: role diversity survives that, model diversity does not.
- Give each seat a depth budget rather than a turn count: shallow for Censor, standard for
Seneca and Librarius, exhaustive for Novator and Scrutator. A host with a turn or step limit maps
these onto it; a host without one just needs the seat to stop when its own output contract is met.
The report says which kind of diversity the run actually got.
Phase 1: Frame
No agents yet. Establish, in the orchestrator's own words:
- The decision — one sentence naming what is actually being chosen. Not the symptom, the choice.
- Candidate A, if an approach is already on the table. A finished plan is not the subject of an
audit here; it enters the board as one candidate, ranked against the others on the same terms.
- Constraints — what genuinely limits the solution space: existing architecture, compatibility,
effort available, things that must keep working.
- Non-goals — what is out of scope, so seats do not solve a larger problem than the one asked.
Where a constraint is a claim about the existing system, check it in the repo rather than
asserting it. A frame built on a constraint that is not actually true wastes every seat on the board,
and this is the only phase where it is cheap to catch.
If the decision cannot be stated in one sentence, say so and stop — the board cannot rank candidates
against a question nobody has written down.
Announce the frame in a few lines. Every seat receives this identical frame; nothing else about the
conversation reaches them.
Phase 2: Diverge
Dispatch all generators at once so they run concurrently. They must not see each other's output.
Do not hint at which candidate you favour, and do not pass the conversation's reasoning about it.
The two native seats read their prompt from
with
replaced by the Phase 1
frame. Dispatch a subagent per seat that returns candidates in the shape its prompt specifies; on a
host with no subagent facility, run the same prompt inline.
- Novator —
references/novator-prompt.md
- Librarius (if selected) —
references/librarius-prompt.md
Peregrinus runs through the collection's external-agent skill,
, in
ask mode.
Invoke it by name rather than reproducing its procedure here — it owns temp-file resolution, run ids,
and the rule that the question is written with a file-write tool and never a shell heredoc. Follow its
ask-mode steps, and pass it four things:
--host <the agent you are>
--preamble <skill-dir>/references/peregrinus-prompt.md
- the Phase 1 frame, and nothing else, as the question
- a timeout of , with the surrounding command timeout set to its maximum. This seat produces
three sections and up to three fully specified candidates; outsider's 300s ask-mode default is not
enough for that, and a timeout here costs the whole leg
Check the preamble path resolved before you dispatch. Consilium is reachable through several
generated symlink trees, so
has to be the directory this
was actually loaded
from. Outsider refuses to run with an unresolvable
and says so — if you see that message,
fix the path rather than dropping the seat, because the alternative is a seat that answers with no
brief at all.
Peregrinus is the one seat with no output contract you control, and the one that saw nothing but the
frame.
Name the agent that actually answered —
prints it on the first line; a candidate
from a board member you cannot identify is not interpretable. Map its Section 2 onto the candidate
shape the others use, and carry its Section 3 — what looks off about the problem as stated — into
Phase 5 as cross-cutting material. That section is the most valuable thing a cold seat produces and it
is not a candidate, so nothing else in the flow would pick it up.
The leg is droppable: with no external agent CLI installed, or with
itself not installed,
the run continues without it. Say so in the report, and say how many generators actually ran — with
Librarius unselected that is
one, and a single-generator board cannot show the design space was
explored. Prefer selecting Librarius in that case even if its trigger is weak.
Phase 3: Assemble the Candidate Set
Before the critics run, merge the generators' output into one numbered set. This is the orchestrator's
job and it is not clerical:
- Include candidate A from the frame, described on the same terms as the rest.
- Merge near-duplicates. Two candidates that differ only in naming or file layout are one
candidate. Keep the clearer description and note both origins.
- Kill the non-candidates. "Use something better" is not a candidate. Anything not concrete
enough to start on is dropped, and the drop is reported.
- Strip attribution, by rewriting rather than by omitting. Critics must not know which seat
proposed what, or which one was already on the table — that is the bias the board exists to
remove. Deleting seat names is not enough: prior-art candidates announce themselves ("adopt
"), and an existing plan reads in the house voice. Restate every candidate in one
common voice at the same level of detail, and order them so the pre-existing approach is not
first. You will still know which is which; the critics must not.
Announce the set as one line per candidate.
Phase 4: Converge
Dispatch all critics at once. Each receives the frame and the full assembled candidate set, and
attacks it comparatively — this board ranks candidates, so an objection that hits every candidate
equally changes nothing about the ranking and must be labelled cross-cutting.
Prompts, with
and
replaced:
- Seneca —
references/seneca-prompt.md
- Scrutator (if selected) —
references/scrutator-prompt.md
- Censor (if selected) —
references/censor-prompt.md
On a host with no subagents, run each prompt in turn and never show one critic another's output. Say
in the report that they were not isolated — a sequential run leaks earlier objections into later ones.
The Objection Contract
Every objection names four things, or it is not an objection:
- Candidate — which one it hits, or
- Condition — the circumstance under which it actually bites
- Bearer — who pays, named from this closed list and no other: , ,
, , . Dedupe and ranking both key on this field, so
free-text bearers make both unstable
- Severity — (rules the candidate out; cannot work, or the cost is unrecoverable),
(candidate survives, trade-off gets worse), (worth knowing, does not move the
ranking)
An objection missing both a condition and a bearer is a preference: reported in its own
section, ranking nothing. Missing one of the two is an incomplete objection, not a preference —
supply the missing half if the candidate text supports it, and drop it if it does not.
requires a named bearer. A
objection without one becomes
, because an
unrecoverable cost nobody bears is not a reason to rule a candidate out.
Phase 5: Consolidate and Verify
Critics over-report, over-rate, and file one insight three times. Cut that down first, then verify —
a board of six with nothing between an opinion and the report is six unchecked opinions.
Consolidate:
- Kill non-objections. No named condition and no named bearer is a preference, not an objection
— move it. An objection whose evidence quotes nothing from the frame or the candidate is an
impression; drop it.
- Kill unproven halves. An objection pairing a demonstrated claim with one nobody could
demonstrate ships as the demonstrated claim alone. The weakest claim sets the credibility of the
whole objection.
- Dedupe. Two critics hitting the same candidate with the same objection is one objection at
the higher severity. Independent corroboration is a strong signal — say so, and never let it look
like two problems.
- Cluster. If one change to a candidate answers several objections, report the root and nest the
rest beneath it.
- Separate cross-cutting from discriminating. Cross-cutting objections belong in the framing
section — they say something about the problem, not about the choice. Peregrinus's Section 3
observations join them here.
- Keep what you killed. Pass the drops from steps 1 and 2 into verification marked .
Verification rules on them too, and a confirmed drop is worth more than an assumed one — the
lenses sometimes find the stated reason for dropping was wrong.
Verify: dispatch the lenses from
references/verification-prompt.md
, all at once, one per lens.
Replace
with the lens name,
with the Phase 1 frame,
with the
assembled set, and
with the consolidated objections plus the drops — a lens told to
quote the candidate and judge against the frame's constraints needs all three in its prompt. Every
lens defaults to refuting what it cannot demonstrate, and each rules only within its own verdict
vocabulary.
| Lens | Question |
|---|
| Is this objection about what the candidate actually proposes, or an invented version of it? Quote the candidate. |
| Under what condition does it bite, and who pays? No condition and no bearer means it is a preference. |
| Can the candidate absorb this cheaply? An objection with a cheap fix is a design note, not a reason to rule a candidate out. |
Merge rule. An objection dies when
shows it attacks something the candidate does not
propose, or when
can establish neither a reachable condition nor an exposed bearer. Those two
lenses are the only ones that refute.
never kills an objection. It demotes one to a
design note on its candidate, along
with the specific adjustment that answers it. A design note does not rank, which means escapability
is the one lens that can keep a
objection from ruling a candidate out — so it must state
the adjustment, and the report must carry it. An adjusted candidate is ranked as adjusted, and the
adjustment is named.
Verification is not a downgrade pass. An objection that arrives reasoned and leaves demonstrated
should come out sharper. A verify phase whose ratings only ever fall is miscalibrated. Verify the
reasoned ones hardest, and anything a critic rated confidently without evidence.
Phase 6: Synthesize
Read
references/synthesis-guide.md
and follow it. It covers ranking the candidates, choosing the
recommendation, when to override the board, and the report format.
Before presenting, judge the board against your own broader context: dismiss what is wrong or
irrelevant, demote what is correct but insignificant, promote what matches a concern you already had,
and note the reasoning for any override. You have context no seat had — use it, and say when you did.
Edge Cases
- No external agent installed — Peregrinus is skipped, the report says so. Never retry.
- itself not installed — a different failure with the same symptom, and a real one:
consilium ships in the review bundle while ships in the assist bundle, so a host with
only one of them installed has a core seat that cannot exist. Say which is missing, name the other
bundle, and run the board without that seat.
- Peregrinus times out — note it and move on. If it timed out at 300s, the timeout was not passed.
- Peregrinus ran unbriefed — if its answer has none of the sections its prompt asks for, the
preamble did not reach it. Discard the output rather than mapping it; an unbriefed answer looks like
a candidate and is not one.
- A seat fails — note it in the report header and continue with what returned.
- Only one candidate survives Phase 3 — valid, and worth saying plainly: report it as a decision
with no live alternative, and say what was rejected and why.
- All candidates carry a Blocking objection — the honest report. Say the frame may be wrong and
hand back the cross-cutting objections rather than picking a least-bad candidate.
- The problem cannot be stated as a decision — stop and say so. Do not invent a decision.
- No objections survive verification — a valid outcome. Report the ranking on trade-offs alone
and say the board found nothing disqualifying.
- A finished plan with no open question — it becomes candidate A and the board still generates
alternatives. If it wins, that is the useful answer.
- A diff was passed instead of a decision — say what this skill is for, point at
, and stop.
- Every candidate came from one generator — say so in the header. A single-generator run cannot
show the design space was explored. Prefer selecting Librarius to avoid it, and re-run with a
different frame if the candidates still feel narrow.
- Only one critic ran — allowed only for small, cheap-to-undo candidates, and the report says the
board had one critic. Otherwise select Scrutator or Censor in Phase 3 and dispatch it.
Rules
- Concurrent dispatch: all generators at once, then all critics at once. Never serialize what can
run in parallel, and never let one seat see another's output within a phase.
- Autonomous: do not ask the user questions mid-run — resolve ambiguity yourself and say how.
- Every objection carries a condition and a bearer, the bearer drawn from the closed list. One
missing both is a preference and ranks nothing; one missing either is incomplete and is completed or
dropped.
- No modifications: this board reads and reasons. It never edits the thing it examines.
- Name the agent that answered for Peregrinus, and say which kind of diversity the run got.
- Honest synthesis: disagree with the board when your broader context warrants it, and say so.