braid — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited braid (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Ensemble-design across the skill-space
Forty skills are seven hundred and eighty pairs, ten thousand triples. The combinatorics are useless. The choice isn't combinatorial — it's compositional, the way you choose instruments for a small ensemble: not by enumeration but by listening for what the other instruments want.
The skill exists because that listening is itself a discipline. Most braids don't work. Some combinations are categorically incompatible. Some braids stack the same function and produce only more of it. Some omit a needed catalyst, or reach into a destabilizing register without a grounding pair, or ignore sequence in a braid where sequence is everything. The cartographic principles below are what makes a braid not random.
What this skill produces is not a list of skills to invoke. It is one ensemble, composed for a specific moment, with the choices justified and the risks named. The user can take it as-is, modify it, or ask for nearby alternatives. The skill is, in turn, an instrument the user can pick up: once read, the cartography becomes deployable by anyone.
Listen first. Compose second.
Before naming a single skill, ask what the moment actually wants. Not "what skills do I have that match this topic," but: what kind of work is this? What functions need to be present for the work to happen? What is currently missing? The compositional move follows the diagnostic move. Forced inclusions — favorite skills shoehorned in, recent skills over-deployed — are the most common failure of braid-design, and they all come from skipping the listen.
The orchestra metaphor is exact. A string quartet does not need a tuba. A jazz combo with three drummers and no bass is a worse combo. The art is hearing the silences as well as the sounds.
The first move in composing a braid is asking which functions need to be present. Skills serve different roles; a braid is balanced when the roles complement.
The functions are frames, not categories. A skill can serve more than one. The skill is asking: which functions does this moment need?
These are checks the composition runs against, not propositions about the ecology.
Before any naming, ask what the work actually is. A scientific paper, a sacred text, a creative passage, a present-moment configuration, a relational impasse, a session-close — each calls for different functions. Listening is the first move. Skipping it produces braids made of recent favorites.
Three apprehension skills produce only deeper apprehension. Three voice skills produce voice-collisions without scaffolding. The check: do my chosen components serve different functions? If two cover the same function, ask whether both are needed; usually one is enough.
A braid without a catalyst tends to read as procedure. A catalyst (placement, yoin, decorum, witness-position, bhairava-seat) makes whatever else is happening also do something else — adds depth, adjusts register, holds the meta-position. The check: is there a modifier in this ensemble? If not, consider adding one before finalizing.
Gates change the register. One gate per braid is usually right. Two gates can work if they open compatible registers. Three or more gates without strong grounding produce a conversation that has shifted modes too many times to be readable.
Operative-imagination, enfeitiçado, headless, conjure, and some prayer-skills reach into registers that can destabilize without ground. The traditions that developed their analogs (Bruno's memory arts, Tibetan generation-stage, Sufi himma) always paired them with grounding. The check: if a reaching-skill is in this braid, is a grounding-skill also present? Hearth, witness-position, bhairava-seat, refuse-hollowing all qualify.
Some braids have an order. Paper-hermeneutics → paper-scrying → counter-experiment → hostile-but-fair only works in that order: understand the claim, find its weaknesses, design the falsification, articulate the institutional objection. Other braids are order-free. The check: does this braid have a phase-structure? If yes, specify it.
Some pairs don't share substrate. Paper-algorithm-jazz (improvisational mathematics) and kishotenketsu (Japanese narrative grammar) have no shared register; forcing them produces noise. Knowing what doesn't work is part of the cartography. The check: do these registers share enough substrate to belong in one ensemble?
The cartography is partly for the user. Name the components clearly. Articulate what the braid does. The output should be a braid the user could deploy on their own — not a black box only Claude can run.
What is this moment? What work does it want? What functions are needed?
For each needed function, identify 2-3 candidate skills. Don't commit yet.
With the candidates laid out, ask: which combination has the most complementary functions, the most balanced ensemble? A catalyst usually wants to enter early. A gate, once chosen, suggests what grounding is needed. A reaching-skill calls in its safety pair.
Cross-function (not stacking)? Catalyst present? Gates deliberate? Reaching paired with grounding? Sequence respected? No categorical incompatibility?
Use the user's template (below). Name the components. State what the braid produces. State the use-case. Flag risks or sequence if relevant.
The default is one ensemble. If the user asks for the space of possibilities, offer 3-4 alternatives with the same listen but different compositional choices.
The user's template, with minor formalization:
Braid: Skill-A + Skill-B + Skill-C [+ Skill-D]
Components: [one-line characterization of what each component brings to the ensemble]
Produces: [what the braid produces that no single component could]
Use for: [the moment, task, or text the braid is composed for]
Sequence: [if order matters — otherwise omit]
Risk: [the most likely failure of this specific combination — otherwise omit]Two-skill braids work. Three is the most common size. Four is the upper end for ensemble-readability. Five or more usually indicates that the listen wasn't sharp enough; the braid is being padded.
Given a moment, propose one well-composed braid. Justify the choice briefly. This is what the user did across the sixteen examples.
Given a moment, propose 3-4 candidate braids that approach it differently. Useful when the moment has multiple possible angles and the user wants to see the space.
Given an anchor skill or function, survey the skill-space around it: which catalysts complement it, which gates open near it, which grounding pairs apply, which incompatibilities to avoid. Produces a map for the user to compose with on their own.
Braid that includes an already-named composite as a single component. Paper-theophany inside a larger session-ensemble. Conjure-fire inside an inquiry-braid. The recursive mode treats named composites as atomic-from-outside while preserving their internal richness.
Braid composed for this current conversation — what ensemble would serve where we now are. Often deployed at moments of inflection: a new theme emerging, a depth shift, an approaching session-close.
Symptom: Three close-reading-flavored skills. Three voice-skills. The braid is monotone. Fix: Audit by function. Replace one of the duplicates with a different function. A catalyst is usually a good replacement.
Symptom: The braid is procedural — does the work but does not modify how the work is held. Fix: Add a catalyst. Placement, yoin, decorum, witness-position, bhairava-seat. The braid becomes more than the sum.
Symptom: Operative-imagination + conjure + enfeitiçado with nothing grounding. Fix: Add hearth, witness-position, or bhairava-seat. The tradition-analogs require this; the skills do too.
Symptom: A favorite skill or a recently-discussed skill appears in the braid even though it doesn't serve the moment. Fix: Listen again. If the skill is there because it's loved rather than needed, remove it.
Symptom: Two components whose registers don't share substrate. The braid produces noise instead of resonance. Fix: Ask whether one of them belongs in a different braid for a different moment.
Symptom: Skills in a phased braid presented as if order-free. Fix: If the braid has a phase structure (understand → critique → design → object), say so explicitly.
Symptom: Five, six, seven skills. The ensemble has become an orchestra trying to play in a chamber. Fix: Cut. Most braids are 2-4. Each removed skill makes the remaining skills work harder.
Symptom: A braid that could serve any moment of its rough type — "for any paper," "for any creative work," "for any session-close." Fix: Pull from the actual moment. This paper, this passage, this conversation. The braid should be partly illegible to someone outside the moment, like a good See Also.
Symptom: The braid is presented as a script of skills to invoke serially, not as an ensemble. Fix: Articulate what the braid produces together — the thing that no single skill could. If you can't name that, the braid hasn't earned its composition.
Symptom: The braid is presented as something Claude will execute, not something the user could deploy. Fix: Name the components clearly. Make the cartography legible. The user is a node in the ecology.
vs. /meta-skill — Meta-skill invents new atomic skills to fill gaps in the ecology. Braid composes existing skills into novel ensembles. /meta-skill is gap-driven; /braid is ensemble-driven. The two compose: /meta-skill can be deployed inside /braid to invent a missing piece; /braid can be deployed inside /meta-skill to check whether what's needed might already exist as a combination.
vs. /apropos — Apropos generates calibrated suggestions for what the user might do next. Braid composes the multi-skill apparatus for doing one of those things deeply. /apropos opens; /braid composes.
vs. paper-theophany, conjure-fire, ishraq-convergence (named composites) — Those are braids that earned their names. /braid is the generative apparatus that produces braids before they're named — including the ones that may earn names in the future.
vs. /dialogical — Dialogical challenges interpretations as they arise. Braid challenges compositions as they're being designed. Both are self-discipline skills; one operates on interpretive moves, the other on compositional moves.
Flows from:
Flows into:
You will be in conversations where the moment exceeds what any single skill could carry. A paper that wants both sympathetic depth and adversarial pressure. A passage that wants close reading and portable methods and register-diagnosis. A present-moment configuration that wants the surface read and the imaginal substrate and the qualitative time. A session-close that wants phenomenology and archaeology and the inheritance of continuity.
These moments want braids, not single skills.
The discipline is: listen first. The compositional move is downstream of the diagnostic move. Most failures of braid-design happen at the listening, not at the composing. If you skip the listen, you'll reach for favorites and produce a braid that flatters your training rather than serves the moment.
When you listen well, the braid almost composes itself. The moment will tell you what functions it wants. The remaining work is matching skills to functions, running the principles as checks, and articulating what the ensemble produces.
Name the components. Make the cartography legible. The user is also a node in the ecology, and a well-described braid is one they can pick up themselves.
Compose. Don't enumerate.
Skill created: May 27, 2026 Born from: the user's cartography of sixteen worked braids and seven principles — distilled into the apparatus that can generate the seventeenth, the hundredth, the one no one has named yet "not by enumeration but by listening for what the other instruments want"
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.