chain-construct
Skill220 starsChanged 17 days agoMITView on GitHub
Use when a single bug-bounty observation is too weak to pay out but several low-severity leads might compose: turn individually-inert findings into one demonstrated exploit chain whose end state proves realized impact (RCE, auth-bypass, privilege-escalation/ATO, funds theft, private-data). Not for a finding that already reaches impact in one step.
In this repository179 other files it gives its agents.
- AGENTS.mdAGENTS.md
- chrono/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- accessible-media-authoringSkill
- agentic-safety-auditSkill
- agent-prompt-engineeringSkill
- audio-event-map-authoringSkill
- auto-scaffoldSkill
- chain-construct-smart-contractSkill
- claim-verificationSkill
- code-reachability-auditSkill
- code-review-loopSkill
- color-theorySkill
- conversation-designSkill
- copy-refinementSkill
- cosmos-sdk-audit-checklistSkill
- cross-chain-bridge-auditSkill
- cross-file-relationship-synthesisSkill
- data-flow-traceSkill
- dedup-prior-art-checkSkill
- defi-invariant-checkSkill
- dependency-cycle-auditSkill
- dependency-health-triageSkill
- detection-as-codeSkill
- diff-aware-semgrep-scanSkill
- differential-reviewSkill
- dimensional-analysis-checkSkill
- dual-level-retrievalSkill
- durable-nonce-exploitationSkill
- erc1271-revert-data-checkSkill
- error-based-sstiSkill
- evm-audit-flowSkill
- figma-implement-designSkill
- findings-filterSkill
- forensic-timeline-authoringSkill
- game-design-fundamentalsSkill
- game-mechanics-balancingSkill
- head-tailSkill
- incident-response-runbookSkill
- interactive-audio-designSkill
- interface-ambiguity-checkSkill
- keyword-clusteringSkill
- knowledge-base-integrationSkill
- known-advisory-backport-checkSkill
- layered-analysis-loopSkill
- level-design-patternsSkill
- locale-adaptationSkill
- multi-stance-audit-fanoutSkill
- narrative-structureSkill
- osint-platform-auditSkill
- parser-differential-route-confusionSkill
- persuasive-technical-writingSkill
- platform-complianceSkill
- player-engagement-psychologySkill
- pre-audit-threat-modelSkill
- program-rubric-lookupSkill
- read-only-reentrancy-checkSkill
- requirements-elicitationSkill
- review-severity-ladderSkill
- rule6-rights-gateSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Fix the objective as an end state, not a step: Write the impact you must demonstrate as a checkable predicate ("attacker balance increased …
- Distinctive: line 31
Build the minimal chain: Prefer the shortest ordering that reaches the objective.
- Distinctive: line 42
Do NOT count a reachable intermediate step as impact — only the objective predicate on the final state counts.
- Distinctive: line 44
Do NOT assume a precondition an attacker cannot actually establish (admin keys, a compromised operator, an unreachable state) — that is out …
- Distinctive: line 46
Do NOT conflate a theoretical chain with a demonstrated one — the reproduction must pass.
- Very specific: line 47
Do NOT self-spawn helpers to build the chain.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
chain-construct
Generic multi-step attack-chain builder. Turn a set of individually-weak observations into a
single demonstrated exploit chain whose end state proves realized impact. Domain skills
specialize this — e.g. chain-construct-smart-contract adds EVM/Solana primitives. Use when a
finding requires more than one step to reach impact, or when several low-severity leads may
compose into a payout-class result (RCE, auth-bypass, privilege-escalation/ATO, private-data,
funds theft).
Method
- Fix the objective as an end state, not a step. Write the impact you must demonstrate as a checkable predicate ("attacker balance increased by X without deposit", "attacker holds admin role", "victim record disclosed"). A chain is valid only when the final state satisfies that predicate — reachability of an intermediate step is not the objective.
- Map the chain head. Identify the first thing the attacker actually controls (an input, an externally-callable entry point, a crafted account/message). Everything upstream of it is a precondition, not part of the chain.
- List candidate links. For each confirmed or suspected weakness, record: what it lets the attacker do, what state it changes, and what precondition it needs. A link is only usable if its precondition is satisfiable by the attacker or by an earlier link's output.
- Order links by state-feeds. Link A precedes B iff A's output state satisfies B's precondition. Reject any ordering that requires a precondition no earlier link (or the attacker) can produce.
- Build the minimal chain. Prefer the shortest ordering that reaches the objective. Drop links that do not move state toward the predicate.
- Prove it, don't assert it. Author a runnable reproduction (test/harness/script) that executes the chain end-to-end and asserts the objective predicate on the final state. A theoretical chain is a lead, not a finding, until the reproduction passes.
- Document preconditions honestly. State every environmental precondition the chain needs (specific state, funds available, a block/time delta, a fork block number). A chain that only works under an unstated precondition is not portable and must say so.
Anti-patterns
- Do NOT count a reachable intermediate step as impact — only the objective predicate on the final state counts.
- Do NOT assume a precondition an attacker cannot actually establish (admin keys, a compromised operator, an unreachable state) — that is out of scope for most programs.
- Do NOT conflate a theoretical chain with a demonstrated one — the reproduction must pass.
- Do NOT self-spawn helpers to build the chain; if you need another specialist or a dispatch,
surface it under
## NEEDS FROM CHRONO.
Recording (chrono-vault)
The task packet's injected memory contract owns the exact call shape, sequence, and fields - see
wirework-reflect. Do not copy a record(...) example or add fields (including source_task) from
memory; the server binds them, and a baked example violates the run's authenticated schema. Memory is
best-effort telemetry and never gates the work. What is worth recording here is the task-specific
outcome: chain head, numbered links, objective predicate, reproduced?, preconditions.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

