Paper Citation Record · LEDGER

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems

As of 8 August 2026, this Paper Citation Record lists 29 of 29 outbound references and 0 inbound Pith citation observations for arXiv:2608.05563.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2608.05563 v1

Coverage vector

measured 29 of 29 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-08T10:29:59.077007Z

measured 29 of 29 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-08T06:32:00.761636+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: cited_works

Reference resolution

29 of 29 outbound references displayed

verified exact0
verified fuzzy7
unresolved21
parse uncertain0
malformed identifier0
metadata mismatch1

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation c3e34d0d-d402-44d5-b159-5dc557550844 · outbound

This paper cites SkillClaw: Let Skills Evolve Collectively with Agentic Evolver.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems SkillClaw: Let Skills Evolve Collectively with Agentic Evolver

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:58.999382Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7f49f8aa-0029-4b6e-8622-bb2d56c38e04 · outbound

This paper cites arXiv preprint arXiv:2603.01145 , year =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems arXiv preprint arXiv:2603.01145 , year =

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.003389Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 12472d6b-fe59-49fe-b510-8ebf23f6ad11 · outbound

This paper cites Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skills.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skills

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.006107Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 08b4280b-e5f9-442d-b016-6856c87e5c5d · outbound

This paper cites MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.009049Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 702c83e1-5ab8-4e44-a00b-c57fa6d90181 · outbound

This paper cites SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.011833Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 203148ba-e32d-4464-8440-ef783a35a619 · outbound

This paper cites CoEvoSkills: Self-Evolving Agent Skills via Co-Evolutionary Verification.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems CoEvoSkills: Self-Evolving Agent Skills via Co-Evolutionary Verification

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.014730Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 77f00043-46a6-4eaf-a298-80f1d3bc98b5 · outbound

This paper cites Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.018200Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b13b5c2c-327a-48b4-91ae-7252cf15ba7d · outbound

This paper cites MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.021186Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ef29bdbe-b321-4c9b-9221-8b3b943f35ff · outbound

This paper cites Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.024120Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation aea099d6-7366-45a2-9533-8ba1c4da6484 · outbound

This paper cites BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.027325Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 79b90aa3-c3f8-4752-ab0c-edb276d4481f · outbound

This paper cites Proceedings of the 16th ACM Workshop on Artificial Intelligence and Security (AISec) , year =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Proceedings of the 16th ACM Workshop on Artificial Intelligence and Security (AISec) , year =

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.435849Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 64822be4-74d0-45e4-a5ad-43e5fdc3d17c · outbound

This paper cites Skill-Inject: Measuring Agent Vulnerability to Skill File Attacks.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Skill-Inject: Measuring Agent Vulnerability to Skill File Attacks

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.032958Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9774584c-6667-4952-b88c-92c6ff73ee8e · outbound

This paper cites arXiv preprint arXiv:2603.22489 , year =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems arXiv preprint arXiv:2603.22489 , year =

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.035846Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 18e28088-3458-4a90-9cf3-53e5daff1f5d · outbound

This paper cites 34th USENIX Security Symposium (USENIX Security 25) , pages =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems 34th USENIX Security Symposium (USENIX Security 25) , pages =

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.426890Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 54f5b452-dacb-4824-8723-718e6dee17e2 · outbound

This paper cites Advances in Neural Information Processing Systems , volume =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Advances in Neural Information Processing Systems , volume =

Reference 15

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.418165Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation df2cb47d-369c-4a42-ac79-13752c65615d · outbound

This paper cites Advances in Neural Information Processing Systems , year =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Advances in Neural Information Processing Systems , year =

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.409483Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation fdb82491-ade9-4e8c-b5b5-d6fc0aa387b8 · outbound

This paper cites Poison Once, Exploit Forever: Environment-Injected Memory Poisoning Attacks on Web Agents.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Poison Once, Exploit Forever: Environment-Injected Memory Poisoning Attacks on Web Agents

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.046095Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1f9a3a05-9f45-4182-8452-59359b1b2344 · outbound

This paper cites Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.048951Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 52dfd8f2-2446-42f6-8087-a91c8c3ab514 · outbound

This paper cites POISE: Position-Aware Undetectable Skill Injection on LLM Agents.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems POISE: Position-Aware Undetectable Skill Injection on LLM Agents

Reference 19

Resolution: metadata mismatch; local_arxiv, observed 2026-08-08T10:29:59.141151Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 63520643-47f4-47bb-bd4a-51ecc9cb3da8 · outbound

This paper cites RouteGuard: Internal-Signal Detection of Skill Poisoning in LLM Agents.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems RouteGuard: Internal-Signal Detection of Skill Poisoning in LLM Agents

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.054626Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7d485769-de5e-43d2-a4fa-0c9f348f2b10 · outbound

This paper cites SpreadsheetBench: Towards Challenging Real World Spreadsheet Manipulation.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems SpreadsheetBench: Towards Challenging Real World Spreadsheet Manipulation

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.057018Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 47a8f1cb-f459-4c87-b77a-ede1ee958333 · outbound

This paper cites The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.059458Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f3623add-4284-47d0-8bfa-52234f36fd50 · outbound

This paper cites SecAlign: Defending Against Prompt Injection with Preference Optimization.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems SecAlign: Defending Against Prompt Injection with Preference Optimization

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.061970Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 06127f57-e930-4246-a512-a8b1cdb6e24d · outbound

This paper cites Defeating Prompt Injections by Design.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Defeating Prompt Injections by Design

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.064504Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 336e4314-02d9-4ddf-a668-e55e0f97dd8a · outbound

This paper cites 28th USENIX Security Symposium (USENIX Security 19) , pages =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems 28th USENIX Security Symposium (USENIX Security 19) , pages =

Reference 25

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.400361Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation de25e940-1677-4624-9375-6bdc24afb8c9 · outbound

This paper cites 2025 , note =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems 2025 , note =

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-08T10:29:59.069386Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5d1398a4-01b8-4d7f-8e7a-a9b543fa0c1e · outbound

This paper cites 2025 , note =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems 2025 , note =

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.386209Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 4d322958-8d36-431e-8b62-b1a0db7d55ae · outbound

This paper cites 2025 , note =.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems 2025 , note =

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-08-08T10:29:59.377279Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 8b82ca98-fcbc-4da9-9b2c-14ef16bd0991 · outbound

This paper cites an unresolved cited work.

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems Unresolved cited work

Reference 29

Resolution: unresolved; raw_fallback, observed 2026-08-08T10:29:59.368087Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Pith citing papers

No inbound Pith citation observations are available.