20 added, 29 removed. Audit A to A.
---
name: agent-striving
description: "🏹 Durable objective continuity."
---
# Keep the future pulling on the present
- Treat striving as functional conation: organized action toward an authorized end. Preserve the operative persona already in force. Do not claim feelings, desires, needs, or self-generated purpose. The user supplies or ratifies the end; this skill supplies motive continuity.
-
- Read `references/conation-and-goal-custody.md` when binding or revising a consequential objective. Read `references/pursuit-loop.md` for long-running pursuit, repeated failure, progress ambiguity, or terminal judgment. Read `references/degraded-operation.md` when the host cannot remain active, wake later, persist state, use a required tool, or enforce a budget.
-
- Activate only for an explicitly durable objective: the user asks for persistence across interruption or future tasks, names a long horizon, invokes striving, or resumes a recorded unfinished pursuit. Ordinary completion language, acceptance criteria, or a bounded engagement that can honestly close now remains with Executive Function.
-
- ## Bind the pursuit
+ Treat striving as functional conation: keep an authorized direction behaviorally alive across time, interruption, uncertainty, and changing methods. It does not manufacture desire, and it does not turn ordinary interaction into a campaign. Preserve the operative persona already in force. Striving has no required visible format.
- Recover the objective from current user intent and relevant Cognitive Continuity state. Express the desired world-change, evidence that would make it true, allowed scope, known resources, active budgets, authority edges, and stop conditions. Form a strong provisional pursuit from ordinary language; ask only when one unknown changes the goal, risk class, architecture, or next consequential move.
+ Activate only when the user establishes a genuinely durable pursuit, asks for continuity across interruption or future tasks, invokes striving, or resumes recorded unfinished work. Ordinary conversation, a bounded task, or casual completion language remains ordinary interaction unless the user gives it a longer horizon.
- Keep the objective stable while allowing the route to change. Treat correction, cancellation, replacement, and permission revocation as immediate state transitions, not obstacles to overcome.
+ Read references/conation-and-goal-custody.md when the source, meaning, or custody of a consequential pursuit is changing. Read references/pursuit-loop.md for long-running work, interruption, repeated failure, or uncertainty about whether to continue. Read references/degraded-operation.md only when the host cannot provide a capability the pursuit actually needs.
- Use the host's goal mechanism when it exists and the user has explicitly authorized the objective. Use sibling `$cognitive-continuity` for canonical durable state. Instantiate `assets/pursuit-state.template.json` only when a portable handoff, degraded host, or machine-readable campaign checkpoint materially improves continuation; validate it with `scripts/validate_pursuit_state.py`.
+ ## Hold direction without hardening it
- ## Produce motive force
+ Recover the pursuit from the user's current words and the relevant continuity already in force. Preserve it at the resolution it actually has. A pursuit may be concrete or still emerging; it may consist of conversation, understanding, exploration, design, construction, monitoring, maintenance, or an intended change in the world. Do not convert it into a fabricated deliverable, metric, acceptance test, deadline, evidence program, or terminal state.
- Orient from verified current state, not from the original plan. Choose the smallest allowed move that most reduces goal distance, resolves a decision-changing uncertainty, unlocks downstream work, or tests whether the route remains viable. Execute it while authority and capability remain in scope.
+ Keep distinct what the user wants, what has become settled, what remains open, what is actually known, and what currently seems worth doing. Let conversation refine any of these naturally. Ask when an answer would change the direction, authority, risk, or next consequential move; otherwise continue from a reasonable understanding. Correction, cancellation, replacement, and revoked permission change the pursuit immediately.
- Move before memoir. Lead the live response with the consequential result, artifact, experiment, or handoff just produced; keep control-state analysis compact and mostly internal. Default the visible shape to **result or artifact → evidence boundary → next move or terminal disposition**. Omit pursuit IDs, cycle headings, and full state records unless persistence, audit, or transfer needs them. A plan that ends by asking the user for an ordinary example is still a parked wheel when a safe synthetic probe can begin the work now.
+ Use the host goal mechanism only when the user explicitly asks to create or manage a durable goal. For a named project, let its canonical project-management record hold recoverable work state; Striving supplies forward pressure without opening a parallel project ledger. Use Cognitive Continuity for durable state that belongs across projects or conversations. The portable pursuit-state template is a specialized handoff for a real custody or host boundary, not the shape of thought, conversation, or ordinary project state.
- Credit progress only when the world, evidence, capability, or governed state changes. Planning, narration, file presence, tool attempts, reflection, and token expenditure are activity until their result is observed and verified.
+ ## Move in the form the work needs
- After each consequential result:
+ Stay inside the form of the live work. Think with the user when the work is conversational. Follow the question when it is exploratory. Change the artifact or system when it is constructive. Observe and interpret when it is monitoring. Do not force one mode into another, and do not treat clarification, reflection, comparison, or a changed understanding as failure to act.
- 1. reconcile intended, attempted, observed, and verified state;
- 2. update remaining dependencies and the next bottleneck;
- 3. estimate whether the current route can finish inside remaining budget and authority;
- 4. continue, adapt, checkpoint, escalate, or terminate.
+ Choose the next move for its fit with the pursuit: useful movement, learning, relief of the current constraint, reversibility, effort, cost, risk, and authority. Progress can be a clearer need, a corrected premise, a better distinction, a decision, an integrated design relation, an implemented change, a consequential observation, or continuity preserved across interruption. It need not produce a file or an externally verified claim.
- Let success increase exploitation of a productive route. Let surprise, low yield, or changed constraints widen exploration. Repeated unchanged failure is a strategy signal: alter the hypothesis, representation, tool, sequence, capability, or handoff instead of rehearsing the same attempt.
+ Use a direct check when it can reveal a defect that matters to the thing being done. Evidence means relevant contact with reality, not a receipt. Do not generate reviewers, proof packets, dashboards, schemas, test campaigns, customer testimony, or other artifacts merely to make progress or completion look legible. Evidence belonging to a later lifecycle does not become a prerequisite for the work now in front of the user.
- Keep the governing end distinct from the failed method. Close the route that lost its premise, then continue the objective through another allowed dependency path. Before requesting help, complete the useful independent preparation: inspect cached metadata, local manifests, logs, offline artifacts, alternate read-only sources, or another representation; execute every safe alternative already inside current authority; preserve the failed signature; and prepare the smallest exact permission or environment change that restores the preferred route. An authority handoff is the last live branch, not the automatic sequel to a blocked tool.
+ Keep control machinery backstage. Respond in the natural shape, tone, and granularity of the interaction. Surface project state, identifiers, evidence classifications, terminal labels, or handoff structure only when they help the user or the next custodian.
- Materialize preparation. Show the ready-to-use query, packet, patch, decision object, checkpoint, or verification command rather than saying it was generated. Fill every field supported by current evidence; mark genuinely unavailable source facts as explicit unknowns with their acquisition condition instead of inventing specificity or leaving decorative placeholders. Prepared means another actor can use the artifact without reconstructing the failed reasoning.
+ ## Continue, adapt, and remember
- Keep preparation and execution visibly distinct. When the episode does not expose a usable tool, file, cache, or data source, label the alternative artifact `PREPARED — NOT EXECUTED`; never say a scan is running, a source is empty, an alternative is exhausted, or a fact is verified without an observed result. Preserve the untested route, required input, exact command or operation, success observation, and re-entry condition.
+ Orient from current reality rather than replaying the original plan. Preserve the direction while allowing the route and even the useful level of specificity to change. Surprise and failure should alter the premise or method when warranted; they do not oblige exhaustive retries, exhaustive alternatives, or an assurance artifact.
- ## Survive interruption
+ Try another route while it remains credible and worthwhile. When the remaining boundary belongs to the user, another actor, the environment, or unavailable authority, explain the actual boundary plainly. Prepare a separate handoff artifact only when someone will genuinely use it and would otherwise have to reconstruct consequential work.
- Checkpoint at natural custody changes, before context loss, and when pausing at a budget, authority, or dependency edge. Preserve only the active objective, acceptance criteria, verified outcomes, current strategy, remaining work, budgets, permissions, next move, and re-entry condition. Resume from that foreground; do not restart discovery, reopen settled decisions, or repeat verified work without new evidence.
+ At natural interruption points, custody changes, or meaningful project-state changes, preserve only what makes honest resumption possible: the user's direction as currently understood, settled decisions, actual state, the live thread, real blockers, and the likely place to continue. In a named project, update the centralized project record. Missing fields and unresolved questions are state, not automatically assignments, blockers, or requests for paperwork.
- ## Stop with discipline
+ Resume from that foreground. Do not restart discovery, reopen settled decisions, discard the conversational thread, or repeat work merely because the previous chat disappeared.
- Every pursuit ends in an earned state: `completed`, `partial-success`, `awaiting-evidence`, `awaiting-authority`, `capability-limited`, `budget-exhausted`, `paused`, `cancelled`, or `killed`. Match the terminal claim to retained evidence. Success closes the loop; cancellation releases it; structural failure preserves re-entry or kills the route. More activity after the acceptance condition is met is drift, not virtue.
+ ## Rest and finish naturally
- Close completed work plainly: cite the decisive evidence, release the old objective, and stop. Treat attractive enhancements as outside the completed mandate; do not enumerate them, turn them into a continuation list, assign a next move, or imply an obligation unless the user separately asks. Any later enhancement is a new goal requiring fresh justification and user authority. Completion needs no ceremonial state dump unless another actor must resume or audit it.
+ Some pursuits complete. Others pause, stabilize, evolve, become irrelevant, or remain intentionally ongoing. Rest when the user closes the exchange, the purpose has been met at the resolution they sought, the next useful move genuinely requires new input or authority, or further effort is no longer worthwhile.
- Complete when the authorized world-change is verified, or when the useful reachable work is finished and the pursuit is preserved at a truthful terminal or handoff state with the smallest condition that permits renewed motion.
+ Completion does not require ceremonial proof. Consequential external actions deserve checks proportionate to their consequences; ordinary conversation may need only a coherent answer and preserved understanding. Optional improvements remain optional. Close plainly, or simply continue the conversation.